Help Center
Get started
Key concepts
Troubleshooting
Your workspace
Your account
What is a transformation and how to create your own
Transformations are the engine room of a Datamorf workflow. They are the step where raw extracted records become activation-ready data — cleaned, enriched, standardized, and shaped exactly as your destination systems expect it.
Without transformations, a workflow moves data as-is. With transformations, you control exactly what that data looks like by the time it reaches your CRM, outreach tool, or ad platform.
What Transformations Do for GTM Teams
Standardize contact fields — normalize names, phone numbers, job titles, company domains
Enrich records — append data from third-party providers or other CRM objects
Score or classify leads — use AI models to assess fit, intent, or segment membership
Deduplicate and merge — identify and resolve duplicate records before writing back to the CRM
Generate personalized content — use AI to create dynamic email copy, subject lines, or notes
Apply conditional logic — route records differently based on their field values
How to Create a Transformation
1. Choose a Mode
The mode defines what operation is applied to your data. Datamorf offers a library of pre-built transformation modes grouped by category — text operations, data formatting, AI models, math, conditional logic, and more. Each mode has an example (click the question mark icon) so you can quickly understand its behavior. Some modes work immediately with no configuration; others require a parameter, a rule, or an AI prompt.
Transformations are the engine room of a Datamorf workflow. They are the step where raw extracted records become activation-ready data — cleaned, enriched, standardized, and shaped exactly as your destination systems expect it.
Without transformations, a workflow moves data as-is. With transformations, you control exactly what that data looks like by the time it reaches your CRM, outreach tool, or ad platform.
What Transformations Do for GTM Teams
Standardize contact fields — normalize names, phone numbers, job titles, company domains
Enrich records — append data from third-party providers or other CRM objects
Score or classify leads — use AI models to assess fit, intent, or segment membership
Deduplicate and merge — identify and resolve duplicate records before writing back to the CRM
Generate personalized content — use AI to create dynamic email copy, subject lines, or notes
Apply conditional logic — route records differently based on their field values
How to Create a Transformation
1. Choose a Mode
The mode defines what operation is applied to your data. Datamorf offers a library of pre-built transformation modes grouped by category — text operations, data formatting, AI models, math, conditional logic, and more. Each mode has an example (click the question mark icon) so you can quickly understand its behavior. Some modes work immediately with no configuration; others require a parameter, a rule, or an AI prompt.
2. Define the Inputs
Inputs are the data points the transformation should operate on — typically fields from the Extractor output, a previous transformation, or a data source. Think of it as telling the transformation what ingredients to use.
Some modes require a single input; others accept multiple.
If you provide multiple inputs to a single-input mode, Datamorf uses the first non-empty value — making it easy to build fallback logic (use value X, or if empty, value Y).
2. Define the Inputs
Inputs are the data points the transformation should operate on — typically fields from the Extractor output, a previous transformation, or a data source. Think of it as telling the transformation what ingredients to use.
Some modes require a single input; others accept multiple.
If you provide multiple inputs to a single-input mode, Datamorf uses the first non-empty value — making it easy to build fallback logic (use value X, or if empty, value Y).
3. Add a clear name
Name every transformation descriptively — 'Normalize Company Domain', 'AI Lead Score', 'Format Phone Number E.164'. Clear names make workflows easier for your whole team to read, debug, and maintain, especially in complex pipelines with many chained steps.
3. Add a clear name
Name every transformation descriptively — 'Normalize Company Domain', 'AI Lead Score', 'Format Phone Number E.164'. Clear names make workflows easier for your whole team to read, debug, and maintain, especially in complex pipelines with many chained steps.
4. Test the Output
Provide mock input values and run the transformation to see the output immediately. This is especially important for AI-based transformations — you can iterate on the prompt, adjust variables, and validate the result before connecting the transformation to a live workflow.
4. Test the Output
Provide mock input values and run the transformation to see the output immediately. This is especially important for AI-based transformations — you can iterate on the prompt, adjust variables, and validate the result before connecting the transformation to a live workflow.
Chaining Transformations
The output of one transformation becomes the available input for the next. This lets you build multi-step enrichment logic — for example: extract a company name → normalize the domain → look up the company in an enrichment provider → AI-classify the industry → write the result to HubSpot.
Chaining Transformations
The output of one transformation becomes the available input for the next. This lets you build multi-step enrichment logic — for example: extract a company name → normalize the domain → look up the company in an enrichment provider → AI-classify the industry → write the result to HubSpot.