Skip to content
Book a demo

Cross-source questions

Snowflake and Salesforce in one question

Ask questions that join Snowflake usage and finance data with Salesforce accounts and pipeline, each read as the person asking, with lineage.

Book a demoHow cross-source answers work

Short answer

One question can span Snowflake and Salesforce, such as which accounts with an open renewal saw product usage fall. Nexatron queries each within the asking person’s access, combines the results in an isolated step inside your workspace, and shows each source’s query. Snowflake and Salesforce are both available.

Example joined questions

Questions to ask. Each needs data from both systems; the answer depends on your own data and definitions.

  • Which accounts with an open renewal opportunity saw product usage in the Snowflake usage tables fall by more than half against their three-month average?
  • What is open pipeline by segment in Salesforce beside last quarter’s net revenue from the Snowflake finance mart?
  • Which customers in the Snowflake billing tables have no Salesforce account owner?

Agree which Snowflake column holds the Salesforce account ID, or which mapping table links the two.

Check before you trust a joined answer

  • Each person’s Snowflake role and Salesforce sharing rules both apply, so two people can get different, correct answers.
  • Note when the Snowflake tables were last loaded: the warehouse can lag Salesforce.
  • Aggregate usage to the account before joining, so many usage rows do not repeat one opportunity’s amount.

How a cross-source answer works

  1. Each system is queried within the person’s access

    Nexatron sends each part of the question to the system that holds it. A system each person signs in to answers as the person asking, and a shared connection reads the same data for everyone enabled on it.

  2. Results are joined inside your workspace

    Nexatron combines each system’s results in an isolated step inside your tenant when the question is asked. Nothing has to be consolidated into a warehouse first, and there is no pipeline to build.

  3. The queries come back with the answer

    The answer carries the generated query for each system and how many records matched, so a reviewer can check which system supplied which figure and repeat the check later.

  4. Joins follow links found at setup

    Nexatron joins on the links setup found between the systems, and says how many records matched. If no record matches, it says so and marks the figures as uncertain.

  5. Currencies are never added up as they are

    Amounts in different currencies are converted with one system’s own exchange rates where a usable rate source exists, and otherwise shown side by side with the reason.

  6. Longer than a single-source answer

    These answers take longer than single-source ones.

More on how it works and its limits: Cross-source analytics and the guide to cross-source analytics without a warehouse migration.