Hours that tie to the paycheck

Overtime lives in weeks, payroll lives in periods, and the two never align: practitioners rebuild that bridge in spreadsheets every cycle. This module stores hours at day grain, so weeks, periods, and lookbacks are all just calendars.

RuleStore hours at day grain with the calendar carrying weeks and pay periods both. Every alignment question becomes a join, not a rebuild.

Neverreport overtime from period totals. Weekly thresholds computed on periods are wrong at every boundary week, silently.

◆ The pagesone model, three reports, gaps that close.

Page	What it answers
The hours star	The model: day-grain time entries, scheduled hours, and the calendar that holds weeks and periods together.
Overtime	Overtime by department and week, and the week-against-period boundary handled by grain instead of guesswork.
Reported against paid	The reconciliation: reported hours to paid hours, with a four-bucket gap that closes exactly.
Missing timesheets	The completeness control the reconciliation cannot be: who was expected to submit and did not, named before the deadline.
Unapproved time	The 148 hours that missed the cutoff, aged by manager, and the daily chase list that empties the queue.
Schedule vs actual	Coverage promised against coverage worked, read beside overtime so a subsidized schedule cannot hide.
The rolling average	Lookback thresholds computed as window functions, with the names on either side of the line.

◆ The three delivered gapsreported consistently in the field, all calendar-shaped.

Gap	Consequence
Weeks against pay periods	Practitioners report no obvious delivered source for overtime by week inside pay periods; boundary weeks get split by hand.
No relative dates	Migrators from dedicated time systems miss yesterday and last-week scheduling for daily reports, practitioner-reported.
Rolling lookbacks by hand	Average-hours thresholds over trailing windows are hand-built each time; delivered reporting has no window functions.

◆ Where the numbers tiethe arithmetic the battery checks.

May 2026 runs on the same population and departments as everything else: the lookback partition sums to the headcount page’s 351, the overtime matrix rows and columns sum, and the reported-to-paid gap decomposes into named buckets that add back exactly. Hours feed the payroll facts the labor cost report prices, one population end to end. Sample values are illustrative, never client data.

Use case

Problem

Overtime, lookbacks, and reported-to-paid checks are rebuilt in spreadsheets because delivered reporting cannot cross weeks, periods, and windows.

What we build

A day-grain hours fact with a calendar carrying weeks and pay periods, plus scheduled hours beside actuals.

What you get

Overtime, reconciliations, and lookbacks as joins on one fact, tied to the same population payroll prices.

How many spreadsheets bridge your weeks to your pay periods?

We ship the day-grain model that makes the bridge a join, in 10 days.

Talk to us

Terms on this page

battery: The automated set of checks, including the crossfoot arithmetic, run on every build of this catalog.
day grain: One row per worker per day. Small enough to rebuild any calendar.
pay period: The payroll cycle weeks cross. The alignment trap’s other half.
boundary week: A week split across two pay periods.
lookback: A trailing window: average hours over the last four weeks.
window function: Database math across neighboring rows. What lookbacks are made of.
reported hours: What workers entered and managers approved.
paid hours: What payroll actually paid. The reconciliation’s other side.