Skip to main content

Building a scope

The scope builder (Data Studio) is the visual canvas where you define exactly which objects and records an operation touches. You'll meet the same builder in four places — Restore, Backup, Archive, and Purge — so everything on this page applies across the platform.

[SCREENSHOT REQUIRED: Data Studio canvas with a root object, parents, and children across levels]

The idea​

A scope is a tree of objects:

  • The root object is the record set you care about — e.g. Accounts created before 2020.
  • Children are objects whose records point at the root (Contacts, Opportunities…). Including a child brings its dependent records along.
  • Parents are objects the root looks up to. Including a parent means lookups reconnect properly on load.

The canvas lays the tree out level by level, and a legend keeps the colors straight: Root, Child records, Parent lookup.

Building step by step​

  1. Add root object — search your source's objects (each shows its record count) and pick one.
  2. Select a node, tick relationships — the inspector on the right lists the object's Parents and Children, each with the linking field and an estimated record cost. Tick to include; Select all / Clear all speed things up.
  3. Filter each object — open a node's Filters dialog: build simple field / operator / value conditions, or write an advanced condition directly. The dialog tells you which dialect applies (Salesforce SOQL when the source is a live org; SQL when the source is a snapshot) and can validate the filter and show how many records match.
  4. Watch the totals — a floating pill shows the scope's object count and record total. On live-org sources, Calculate exact runs real counts per object so each card shows what its filter actually matches.

Restore-only tools​

When the builder runs inside a restore, each node also offers:

  • Anonymize — mask sensitive fields with generated values on load. See Data mapping and anonymization.
  • Ignore fields — fields that are never written to the target org (formulas, org-specific, or sensitive fields).
  • Match by external ID — upsert this object by one or more fields instead of the job's load mode, making re-runs idempotent and cross-org loads reliable.
  • Options — per-object load options, such as including the records' files and attachments.
  • Data mapping (toolbar) — graph-wide value rewriting rules.

Good to know​

  • Cards can be collapsed, expanded, and centered; drag the empty canvas to pan.
  • Removing a node removes everything beneath it — the menu shows how many objects that is before you confirm.
  • Leaving the builder with unsaved changes prompts a warning first.