Skip to main content
A census is the list of people you send out for quoting. You need enough on each of them to get a price: how old they are, where they live, and who they would cover. Anything past that is data you did not need to share.

​
What you’ll use

Models - employee, dependent, location Events - connector_synced
Before you start
  • You know which fields the quote needs. Changing scope later means a resync.
  • You’ve agreed with the customer which employees are being quoted, including anyone on leave or retired.
  • The connector has synced. A census read mid-sync is quietly short, so gate the extract on connector_synced rather than a timer - see Triggering a job on completion.

​
Steps

1

Confirm the scope carries every rating factor

Read one employee and check that date_of_birth, home_location, gender and tobacco_use come back populated. Map marital_status too, as a cross-check against the dependent roster.Scope out ssn, bank info and ethnicity as well, since a quote prices on none of them. Widening scope later triggers a resync, so settle it now - see Scoping.Result: Every rating factor confirmed present.
2

Fetch the population

Expand home_location. The rate follows where someone lives, while work location sets your customer’s situs.
Filter to the population you agreed with the customer. employment_status=ACTIVE on its own drops people who still belong on the census: LEAVE is still enrolled and still rated, RETIRED is often quoted as its own population, and PENDING covers new hires inside the waiting period.Where your customer spans more than one legal entity, filter on company_id. Full-time and part-time split on employment_type, which sits on employments rather than the employee record.Follow the cursor to the end - see Pagination. A census that stops at page one looks like a success.Result: The population being quoted.
3

Add dependents

relationship is what builds the tier - employee only, employee plus spouse, employee plus children, or family. Date of birth alone cannot tell a 24-year-old spouse from a 24-year-old child. is_student keeps a child eligible past the plan’s age limit.Send the eligible roster rather than current enrollment, or the population comes out short. DOMESTIC_PARTNER, SIBLING and PARENT are all valid values, and not every carrier accepts them as dependents - confirm before mapping them into a tier.Result: The dependents each employee could cover.

​
Frequently Asked Questions

Reconcile against the connector’s own counts before re-reading - see Record counts.
Either scope excluded it, or the customer’s credentials cannot read it - see Scoping.
  • Filtering - how filters, expand and modified_after narrow a read
  • Record counts - when the headcount doesn’t match the customer’s
  • Benefits - the four benefit models, and how dependents are counted