Skip to content

Configuration reference — what ships and what you build

The most common early misunderstanding is assuming a fresh install arrives configured. Some of it does. A great deal does not, and the two are easy to tell apart once you know the rule.

The rule

Ships with content? Why
Custom Metadata Types Yes Packageable — the vendor's defaults travel with the package
Custom object records No Data, not metadata — cannot be packaged

So scoring weights and field mappings arrive working, and trigger settings arrive populated but switched off (see Turning automation off). Rollup definitions, governance policies, and classification assignment rules arrive empty, because they encode your firm's policy rather than the vendor's.

What is empty in a fresh install (checked on 1.1.7)

Object Rows on install
Rollup Definitions 0
Governance Policies 0
Data Quality Issues 0
Classification Assignment Rules 0
ClientVerse Logs 0

Empty is correct for all of these — but classification assignment rules being empty means nothing is classified automatically until you write some. That surprises firms who assume tagging is automatic.

Custom Metadata Types

Core 1.1.10 ships 43 packaged Custom Metadata Types, and Modules 0.8.0 adds 15 more (see catalog.json). They are the configuration surface. A selection of the ones you are most likely to touch, with row counts taken on 1.1.7:

Type Rows shipped Controls
Quality_Score_Weight 17 How data-quality scores are composed
Collaborator_Field_Mappings 9 Which objects participate in wall sharing
Classification_Picker_Mapping 8 What the Classification Picker offers per object
Relationship_Coverage_Rule 7 Relationship coverage evaluation
Engagement_Score_Weight 6 How engagement is scored
CV_SeedDataConfig 4 Seed-data generation
CV_Named_Query 4 Registered safe queries
Client_Status_Config 3 Client status derivation
Quality_Threshold 3 Quality bands for Account, Contact, Opportunity
CV_Trigger_Setting 3 Trigger enablement
Rollup_Config 1 Rollup behaviour
Object_Naming_Rule 0 Dynamic record naming — yours to define

Edit them at Setup → Custom Metadata Types → the type → Manage Records.

Changes to these are org configuration and survive upgrades. An upgrade may add rows; it will not silently revert yours.

Engagement scoring — the actual numbers

The User Guide tells users to ask their administrator what Engagement Score means. Here is the answer, and it is configurable.

Engagement_Score_Weight ships six active signals:

Signal Points per occurrence
Campaign Responded +20
Meeting Completed +10
Active List Membership +5
Content Send +5
Email Exchanged +5
Recency Decay Per 30 Days −5

Each row has Signal_Name__c, Points_Per_Occurrence__c, and Is_Active__c.

The weighting encodes a point of view, and it is worth checking against your firm's. A campaign response counts four times a meeting, on the reasoning that a response is an unprompted signal of interest while a meeting may have been the firm's initiative. If your firm disagrees, change the numbers — but change them deliberately and once, because the score is not comparable across a change.

Recency Decay is what makes the score behave like a relationship rather than a tally. Without a negative term, a contact engaged heavily three years ago outranks one engaged modestly last month. If you disable it, Engagement Score stops meaning "currently engaged" and starts meaning "engaged at some point" — and the Previous Engagement Level comparison that users rely on becomes far less useful.

Data quality scoring

Quality_Score_Weight (17 rows) composes the score; Quality_Threshold (3 rows) sets the bands for Account, Contact, and Opportunity.

Review the thresholds before switching data-quality reporting on. Bands that are too strict produce a dashboard everyone learns to ignore, which is worse than no dashboard.

Turning automation off

CV_Trigger_Setting provides a kill switch through a single Triggers_Enabled__c field, with three rows shipped: Org_Default, CampaignMember and Search_Component_Config.

This is what you use during a large data load, when trigger execution would be both slow and unwanted. Set it back afterwards — and note that records loaded while triggers were off did not get the derived values the triggers produce, so plan a recalculation.

In 1.1.10, CV automation ships switched off. Org_Default ships with Triggers_Enabled__c set to false, so the features built on it do nothing until it is enabled for each object: dynamic record naming, list-member sponsor attribution, the pitch lock, 🧩 email intake, and walls and collaborators.

Rollups

Rollup Definitions ship empty. Rollup_Config supplies one row of behaviour; the definitions themselves are yours.

Rollup Definition records carry a Recalculate action for re-running a rollup after a bulk change or a definition edit.

Modules configuration

With ClientVerse Modules installed, further types appear: Dispatch_Config and Dispatch_Role_Mapping for digests and notifications; Whitespace_Score_Weight, Whitespace_Scoring_Config, Whitespace_Threshold and Whitespace_Tier_Score for opportunity scoring; CV_AI_Action_Template, System_Prompt_Config and AI_Adapter_Limit for the AI features, to which 0.8.0 adds Email_Intake_Setting, AI_Cleansing_Setting, AI_Cleansing_Action, AI_Cleansing_Threshold and CV_Semantic_Field; and CV_BI_Connection for Analytics. Key_Vault is a Core type, so it is present without Modules.

The Whitespace types mirror the Engagement pattern — weights plus thresholds — and the same advice applies: check the weighting against your firm's view before trusting the output.

What to configure, in order

For a new implementation:

  1. Org-wide defaults and sharing — everything else assumes this. See Who sees what
  2. Permission set group assignments — so people can log in and see something
  3. Post-install steps — buttons, flow activation, page assignments
  4. Classification taxonomy — before importing data that should be tagged
  5. Classification assignment rules — so tagging does not stay manual forever
  6. Rollup definitions — once the data is in
  7. Quality thresholds and engagement weights — review the defaults, adjust deliberately
  8. Modules configuration — if licensed

Steps 4 and 5 are the ones most often deferred and most expensive to retrofit: tagging data after import means revisiting every record, while tagging at import costs nothing.