The history window that closes behind you

Workday's delivered workforce history lives in the Trended Worker data source: month-end snapshots, kept for a rolling window, then purged. Useful, and narrower than most teams assume until the month they need what it dropped.

RuleTreat the trending store as a cache with a horizon. Copy every snapshot into your own warehouse before it ages past the window.

Neverpromise a trend longer than the window from delivered data alone. The months beyond it are not slow to fetch, they are gone.

◆ What the trending store isthe plain version, before the limits.

Most Workday reporting reads the live tenant: what is true right now, with effective dating to look backward one date at a time. The Trended Worker data source is the exception. On a schedule, Workday writes a snapshot of the workforce as of the last day of each month and keeps a rolling set of them, so period-over-period reports have something to stand on.

That design is sensible and honest about what it is: a bounded cache for recent trend reporting. The trouble starts when it is mistaken for an archive. Three limits do the damage, and all three are structural rather than configuration mistakes.

◆ The three limitsgrain, window, and what a reorganization does.

Limit	What it means	What it breaks
Monthly grain only	Snapshots are taken as of the last day of each month. There is no weekly or daily option to configure.	Any question about a mid-month date, a week-end census, or the day before a restructuring. Exit dates rarely land on month ends, so timing mismatches creep into rates.
A window of about 36 periods	Roughly three years of month ends are retained, and snapshots older than the window are purged. Both figures practitioner-observed; retention has shifted without much notice.	Five-year trends, cycle-over-cycle comparisons, and any audit question about a period that has aged out.
History is not reorg-proof	Snapshots record the organization as it stood, and the delivered rollups follow today's hierarchy.	Comparing a department across a restructuring. The old rows exist, but they no longer roll up to anything comparable.

Workday's own workaround is to buy its analytics add-on and load snapshots into it before they age out, which practitioners then describe as gappy and fragile, and which still lives inside the vendor's walls at the vendor's price. The structural fix is smaller and duller: keep your own copies, in your own warehouse, on your own schedule.

◆ The owned snapshot beside itsame idea, no window, your grain.

The owned version of the same idea is a daily snapshot fact written by the extraction pipeline: one row per worker per day, retained indefinitely, rolled up through a versioned organization dimension so old months stay comparable across restructurings. The headcount as of any date page shows the table and the one-filter query; the turnover calculation page shows what honest denominators do with it.

-- month-end history, delivered window vs owned table
-- delivered: about 36 rows per worker, monthly, then purged
-- owned: every day since go-live, one filter away from any grain
SELECT snapshot_date, COUNT(*) AS headcount
FROM fct_worker_day
WHERE is_active
  AND snapshot_date = LAST_DAY(snapshot_date)
GROUP BY 1
ORDER BY 1

The owned table subsumes the delivered one: filter to month ends and you have everything the trending store held, minus the horizon. The cost is a pipeline you were building anyway for the Core HR star, plus storage measured in coffee money.

Use case

Problem

Workforce trends are built on the delivered trending store, and every request beyond three years, below monthly grain, or across a reorganization dead-ends into a manual rebuild from old exports.

What we build

A daily worker snapshot in your warehouse with a versioned organization dimension, filled forward from go-live, so the delivered window stops being your history's ceiling.

What you get

Every trend the trending store can produce, plus the ones it cannot: any grain, any date, any number of years, comparable across restructurings.

Still inside the 36-month window? Good, there is time to copy it out.

We ship the snapshot pipeline before the oldest months age away.

Talk to us

Terms on this page

HR: Human resources. Core HR is the module of record for people data.
Trended Worker: Workday's delivered data source of month-end workforce snapshots.
data source: The population a Workday report is built on. One per report.
snapshot: A dated copy of a table's state, kept so the past stays queryable.
retention window: How long data is kept before being purged. Here, about 36 monthly periods.
grain: What one row means. Monthly in the delivered store, daily in the owned one.
versioned dimension: A dimension that keeps each historical value with its dates, so old rows roll up as they did then.
practitioner-observed: A limit seen consistently in the field but not published by the vendor. Labeled until tenant-verified.