The Problem: Ribbon-Cutting Day, Everything Looked Perfect

New warehouse, ribbon cutting, forklifts lined up in formation, the first pallet received without a hitch — the photos got plenty of likes. Then week 3: WMS and ERP inventory won't reconcile. Week 5: pickers call the location coding "inhuman." Week 8: the first major client audits the site and finds pallets blocking the fire lane — the corrective-action notice lands on the desk.

Warehouse launches rarely die on "day one." They die in the 30-to-100-day stretch of systematic exposure: processes never stabilized, staffing never completed, issues never closed. This 100-day checklist breaks zero-to-one into four phases, each with explicit deliverables and red lines.

The Method: Four Phases, One Deliverable Each

Phase 1: Days 1–15 — foundation (a day late here, three months of chaos later)

| # | Checkpoint | Done-when | Red line | |---|---|---|---| | 1 | Location coding scheme frozen | Every location labeled; coding standard signed off | No changes for 90 days after go-live | | 2 | WMS master data | SKU masters, location masters, user roles imported | Never "go live now, backfill data later" | | 3 | Network & hardware | Full Wi-Fi coverage tested; handhelds/printers/scanners on site | Zero dead zones before opening | | 4 | Safety systems | Fire inspection passed, forklift licenses verified, emergency plan posted | No fire sign-off, no operations |

Location coding is the foundation of the foundation. We have seen a warehouse change its coding scheme two months after opening: full relabeling, WMS data migration, retraining — three days of shutdown. Design the code as zone–aisle–level–position, and leave expansion room (3-digit aisle numbers even if you only have 20 aisles today).

Phase 2: Days 16–45 — pilot (run every process 3× on small volume)

This phase takes one or two small clients, or small-batch in-house volume only. The goal isn't output — it's running every process at least three times. Required deliverables:

  1. Signed-off SOPs: receiving, putaway, picking, audit, packing, shipping, returns, cycle counting — eight core SOPs, each validated and signed by the frontline workers who run it, not written in an office.
  2. Exception playbook: every exception encountered in the pilot (no-ASN arrival, unscannable barcode, full location…) gets an owner and a resolution SLA. Target: by day 45, 90% of exception types have a standard response.
  3. Integration test report: WMS ↔ ERP, WMS ↔ carrier/TMS, WMS ↔ client EDI — every interface runs at least 50 real transactions with variances at zero before moving on.

Red line: system bugs found in the pilot must be fixed before volume ramps. Ramping with known bugs multiplies pilot problems by 100. We watched one warehouse ramp early for peak season — a WMS wave bug mis-shipped 2,000 orders and cost $400,000.

Phase 3: Days 46–75 — ramp (climb the capacity ladder, don't open the floodgates)

| Week | Volume target | Staffing | Watch metric | |---|---|---|---| | Week 7 | 40% of design capacity | Permanent staff + 20% temps | Error rate < 0.8% | | Weeks 8–9 | 70% of design capacity | Permanent staff + 40% temps | UPH at 80% of P50 line | | Weeks 10–11 | 100% of design capacity | Full roster | Error rate < 0.5%, UPH on target |

During ramp, hold a 15-minute daily standup covering three numbers only: yesterday's error rate, UPH attainment, open exceptions. If open exceptions rise three days straight, freeze the ramp immediately — the most important brake pedal of this phase.

Also complete client-audit readiness in this phase: get major clients' audit checklists 30 days early and self-inspect line by line. Fire lanes, temperature logs, pest control, training records — the big four of every audit, all mandatory.

Phase 4: Days 76–100 — lock in (turn "tribal knowledge" into "rule of law")

  1. KPI system goes live: set P50/P75 lines from the first 75 days of data (see "Warehouse KPI Systems" in this series); formal evaluation starts day 76.
  2. Cycle counting launches: ABC tiers + counting frequencies running by day 80 (see "Cycle Counting vs. Wall-to-Wall").
  3. Retrospective + handover: on day 100, hold the formal launch retrospective, publish the open-issue list with closure status, transfer unresolved items to daily operations, and disband the launch team as operations takes over.

Red line: by day 100, SOP deviation must be under 10% (sample 100 task executions; the share performed per SOP). If the floor still runs on experience, the SOPs are wall decorations and the launch isn't done.

Field Case: A 3PL Greenfield's 100 Days

Workers installing racking beams in the new warehouse

Background: a 120,000 sq ft greenfield 3PL site in Texas, planned for 3 e-commerce clients, design capacity 15,000 orders/day. The owner wanted "full volume in 60 days."

What actually happened (we talked him into the checklist): Days 1–15, foundation — found 6 Wi-Fi dead zones, 5 days to fix (unchecked, handheld dropouts would have driven everyone insane). Days 16–45, one small client only (800 orders/day), surfacing 47 process bugs — all fixed. Days 46–75, three-week ramp to 15,000 orders; week 9 error rate touched 0.9%, ramp frozen for 4 days, root cause traced to new hires skipping pick verification, retraining fixed it. Days 76–100, KPIs live, cycle counting launched, day-100 retrospective with 3 open items transferred to operations.

Outcome at day 100: 15,200 orders/day, 0.42% error rate, 91% UPH attainment, all 3 clients passed their audits. The owner's words: "We were 40 days 'slow' and saved at least $2 million in rework and claims. Those 40 days were the best 'slow money' I ever spent."

Pitfalls to Avoid

  1. Don't compress the foundation phase for a ribbon-cutting date. Location coding, WMS master data, network coverage — open without these and the next 100 days are debt repayment. The ceremony can wait a week; the debt cannot.
  2. Overlap the launch and operations teams 30 days early. Before the launch team disbands, the open-issue list must be walked through with the operations manager face to face, signed off. The classic failed ending: "the project team left, and nobody owns the problems."
  3. Get the first big client's audit standard before signing. Many 3PLs learn the audit requirements after the client moves in — remediation then costs 5×. Ask for the audit checklist during contract negotiation.
  4. Pre-authorize the "freeze the ramp" decision. The ramp-phase brake must be delegated to the operations manager before launch, no escalation needed. By the time approvals come back, 2,000 orders have shipped.

Summary

One hundred days, four phase keywords: foundation (15 days: coding/data/network/safety), pilot (30 days: small volume, 3× through every process), ramp (30 days: stepped climb + brake pedal), lock-in (25 days: KPIs + counting + retrospective). Every phase has explicit deliverables and red lines — and red lines are not to be crossed. A launch isn't a 100-meter sprint; it's a 100-day marathon. The first 15 days of foundation decide whether the next 85 are highway or mud. Print this checklist, pin it on the war-room wall, cross off one item a day — 100 days from now you'll thank yourself.