joecastelino

amg-wip-payroll-vs-rth-analysis

Analyze an AMG per-store semi-monthly WIP workbook (tech payroll vs RTH / labor relieved to WIP) and produce Joe's service-manager meeting prep — trend table, per-tech gap ranking, talking points, recoverable-$ target. First run 2026-07-08 on "2026- SCWV Wip.xlsx" (Stevens Creek VW); Joe approved the format verbatim.

joecastelino 2 1 Updated 1w ago
GitHub

Install

npx skillscat add joecastelino/jay-skill-pack/amg-wip-payroll-vs-rth-analysis

Install via the SkillsCat registry.

SKILL.md

AMG WIP Payroll-vs-RTH Analysis (Service Manager Meeting Prep)

When to use

Joe drops a store WIP workbook (e.g. "2026- SCWV Wip.xlsx") and asks for "input" ahead of a meeting with the store's service manager. NOTE: this is DISTINCT from /home/itadmin/amg-wip/AMG-WIP.xlsx (the monthly fixed-ops metric tracker, rows=metrics cols=months). This one is the payroll reconciliation workbook.

Workbook structure

  • One tab per SEMI-MONTHLY pay period (two per month, e.g. "1-15" and "16-31" halves), Jan onward.
  • Rows = technicians. Columns include gross pay, pay after R&R/OT adjustments, OT premium, RTH (retail time hours / labor $ relieved to WIP), and the gap allocation.
  • The payroll-minus-RTH gap is booked to GL accounts:
    • 5416 Unapplied Labor — the number that matters
    • 6424 Sal & Wages — guarantee/hourly techs (lube/shop/detail roles)
    • 6444 Training
  • Also cross-references monthly service gross.

Analysis recipe (Joe-approved format)

  1. Parse with openpyxl/pandas, combine both halves per month for full-month figures.
  2. 6-month trend table: Month | Payroll | RTH relieved | Applied % (RTH/Payroll) | Unapplied booked. Bold the best and worst months.
  3. Per-tech table for the problem month: Tech | Pay (after R&R/OT) | RTH | Applied % | $ Gap, ranked by gap descending. Tag which techs' gaps flow to 6424 (guarantee). Call out best performers too (shows it's achievable).
  4. Key diagnostic framings Joe uses:
    • Half the shop under 50% applied = shop-loading/dispatch problem, not one bad tech.
    • Best recent month proves the roster CAN hit that applied % → recoverable $ = (best% − actual%) × payroll. Anchor the meeting on that number.
    • Guarantee bucket (6424) = "do these guarantees still make sense at current volume?"
    • OT premium paid while unapplied is high = OT-approval process question.
    • Unapplied as % of service gross for scale.
  5. Talking points, not a to-do list: per Joe's meeting-prep rule (see amg-store-manager-meeting-prep skill), frame as questions Joe asks + levers JOE owns, not orders for the manager.
  6. The killer cross-check offer: pull the store's Tekion Tech Performance (Beta) report (/core/reports/service/tech-performance — attendance hrs vs flagged hrs, proficiency %) for the same period. If proficiency is fine but applied % dropped → hours exist but aren't being relieved to WIP correctly (flagging problem); if proficiency also dropped → dispatch/traffic problem. This answers "is it dispatch or is it flagging" before the meeting.

Baseline (SCVW, H1 2026)

Applied % ranged 53.4% (Jun, worst — $31.8K unapplied) to 69.5% (Apr, best — $14.7K). Monthly payroll ~$95–110K. Use April-level ~70% as the achievable target at SCVW.

Delivery

Email through Stacey (she owns report emails) to Joe's inbox, HTML tables, worst numbers flagged red. Remember Gmail self-send dedup: From==To==Joe lands only in Sent — Stacey must also append a copy to INBOX; verify in INBOX, never Sent Mail.

Pitfalls

  • Semi-monthly halves must be COMBINED before month-over-month comparison — a single half looks artificially bad/good.
  • Training (6444) coding is inconsistent across periods — mention but don't build conclusions on it.
  • Use pay AFTER R&R/OT adjustments for the per-tech gap, not raw gross.