Salesforce Dictionary - Free Salesforce GlossarySalesforce Dictionary
DictionaryRRecipe
AnalyticsIntermediate

Recipe

In CRM Analytics (formerly Tableau CRM, originally Einstein Analytics), a Recipe is a set of data transformation steps that prepares and combines datasets for analysis.

§ 01

Definition

In CRM Analytics (formerly Tableau CRM, originally Einstein Analytics), a Recipe is a set of data transformation steps that prepares and combines datasets for analysis. Joins, filters, computed columns, and aggregations are defined in a visual editor, each step connected to the next to form a declarative data-prep pipeline - the output is a dataset that dashboards and lenses read from.

§ 02

In plain English

👋 Study buddy

A Recipe in CRM Analytics is a visual data pipeline. Drag in your sources, add steps to join, filter, compute, and aggregate, and the output is a clean dataset your dashboards use. Same tool was called Tableau CRM Recipe and Einstein Analytics Recipe in earlier years.

§ 03

Worked example

scenario · real-world use

An analytics engineer at Harbor Logistics builds a Recipe for the monthly shipment dashboard: it ingests Shipment records from Salesforce, joins to a Carrier dataset uploaded from a spreadsheet, filters to the trailing 12 months, computes an on-time percentage per lane, aggregates to carrier-and-region level, and outputs a dataset. The recipe runs on a nightly schedule; the monthly dashboard reads the clean dataset instead of re-joining raw data each time someone opens it.

§ 04

Why Recipe matters

In CRM Analytics (renamed from Tableau CRM in 2022; originally launched as Einstein Analytics), a Recipe is a set of data transformation steps that prepare and combine datasets for analysis. Steps are defined in a visual editor - input, join, filter, transform, aggregate, output - and connected to form a declarative data pipeline. Recipes are how analysts shape raw source data into the form that dashboards and lenses can efficiently query.

Recipes are foundational to CRM Analytics quality because dataset shape determines what analysis is possible. Well-designed recipes produce clean, appropriately denormalized datasets that make dashboards fast and flexible; poorly designed recipes produce data that every downstream dashboard has to work around. Mature analytics teams treat recipe design as first-class data engineering - naming conventions, documentation, and code-review discipline make recipes maintainable as the library grows.

§ 05

How to create Recipe

Recipes are the data-prep workflows in CRM Analytics — visual transformations that combine datasets, apply filters, compute columns, run aggregations, output a final dataset. Replacement for the older Dataflow tool; modern CRM Analytics development uses Recipes for all data prep. Output is a dataset that lenses / dashboards / Einstein Discovery models read from.

  1. Open CRM Analytics Studio → Recipes (or App Launcher)

    App Launcher → Analytics Studio → Create → Recipe.

  2. Set Recipe Name and target App

    Convention: per-purpose ("Pipeline by Stage," "Customer 360").

  3. Add input nodes (datasets / Salesforce objects / external sources)

    Drag onto canvas. Each input is a starting dataset.

  4. Add transformation nodes

    Join (merge two inputs), Filter (rows), Transform (compute columns), Aggregate (summarize), Append (union). Connect via arrows.

  5. Configure each node's parameters

    Click into a node → set field mappings, filter conditions, formula expressions.

  6. Add an Output node

    Specify the target dataset — what name, which App, which level of permissions.

  7. Run the Recipe

    Click Run to execute the pipeline. First run creates the dataset; subsequent runs refresh it.

  8. Schedule recurring runs (optional)

    Recipes can run on schedule (hourly / daily / weekly) to keep the output dataset fresh.

Gotchas
  • Recipes replaced Dataflows for new CRM Analytics development. Older orgs may have both — migrating Dataflows to Recipes is a manual rebuild.
  • Heavy Recipes can take long to run. Joins across multi-million-row datasets can take 30+ minutes — schedule for off-peak times.
  • Output dataset permissions are configured at the output level. Without proper Row-Level Security predicates, the dataset may expose data the source restricted — audit before publishing.
§ 06

How organizations use Recipe

Apex Analytics

Maintains a library of production Recipes for common prep patterns (cohort builds, hierarchy flattening, period-over-period joins) so every new dashboard starts from a shared dataset.

MarketPulse

Uses Recipes to combine Salesforce data with external sources (web analytics, ad platform exports) into unified datasets that power cross-channel attribution.

SilverLine Corp

Applies engineering discipline to Recipe design - naming standards, description fields, change review - so analysts can safely modify each other's pipelines.

§

Trust & references

Keep learning

Hands-on resources to go deeper on Recipe.

Was this entry helpful?
Help us write better definitions. Quick reactions or detailed edit suggestions.
§

Test your knowledge

Q1. What is a Recipe in Tableau CRM?

Q2. What operations can recipes do?

Q3. Why invest in recipe quality?

§

Discussion

Loading…

Loading discussion…