A spreadsheet is a powerful, free tool that most engineers already know. It can produce a WPS that an auditor will accept. The question isn't whether it's possible — it's whether it's the lowest-cost path when you account for everything.
Here is the honest accounting.
Direct cost comparison
Excel: $0 incremental cost. Comes with Microsoft 365 most shops already have.
Desktop incumbent (WeldAssistant, WeldOffice, etc.): $1,500–$5,000 per seat one-time, $300–$800/year support. For a 3-seat license: $5,000-$17,000 over five years.
Cloud-based modern tool: $99–$499/month per organization. For a small shop on a $249/month plan: ~$15,000 over five years.
Excel wins on direct cost. By a lot. So why isn't every shop using Excel?
Indirect cost — the audit-finding tax
Audit findings cost money. Estimate ranges:
- Minor finding (single WPS deficiency): $500–$1,500 to remediate
- Moderate finding (revision-control breakdown): $2,000–$5,000
- Major finding (procedure not qualified): $5,000–$25,000
- AISC certification at risk: $25,000+
A spreadsheet-based WPS library typically produces 5-15 findings per annual audit. That's $5,000–$50,000+ per year in remediation costs, ignoring opportunity cost.
A software-based library with rule engine validation typically produces 0–3 findings per audit. That's $0–$5,000 per year.
The annual delta is $5,000–$45,000.
Indirect cost — authoring time
The clock you don't see:
- Hand-written WPS in Excel: 60–90 minutes
- Software with rule engine: 5–15 minutes
For a shop drafting 30 WPSs per year (new + revisions), that's 30 × 75 minutes = 37.5 hours for Excel, vs 30 × 10 minutes = 5 hours for software. At a $50/hour loaded rate, that's $1,875 vs $250 — annual delta $1,625.
Indirect cost — audit-day preparation
When the auditor arrives:
- Excel library: 2–4 days of preparation (pulling files, verifying revisions, organizing PQR archive)
- Software library: 30 minutes to one hour (one-click export of current revisions plus the audit packet)
For one annual audit, Excel costs 3 days × 8 hours × $50/hour = $1,200 in prep time. Software: $50. Annual delta: $1,150.
Indirect cost — onboarding new CWIs
- Excel library: 2–4 weeks for a new CWI to learn the shop's filing conventions
- Software library: 2–3 days
At $80/hour loaded rate for a CWI, the difference is significant during turnover.
The numbers, summarized
For a small fab shop with ~30 WPSs and an annual audit:
| Cost category | Excel | Software |
|---|---|---|
| Direct software cost | $0 | $3,000/year |
| Audit-finding remediation | $5,000–$45,000 | $0–$5,000 |
| Authoring time (annual) | $1,875 | $250 |
| Audit prep (annual) | $1,150 | $50 |
| Annual total | $8,025–$48,025 | $3,300–$8,300 |
The break-even is typically reached in the first year, often in the first audit cycle.
When Excel actually makes sense
A few shops where Excel is the right call:
- Very small shops with 5 or fewer WPSs, no certifications, no annual audits. The overhead of software isn't justified.
- Air-gapped facilities with no internet access (some defense contractors). Some desktop tools work; cloud tools don't.
- Shops in transition where the WPS library is being deprecated and replaced. No new investment.
Everyone else is paying the Excel tax whether they realize it or not.
How to make the move
A six-step migration from Excel to software:
- Export every Excel WPS to PDF as a frozen archive
- Re-enter parameters into the new tool, current revision only
- Sign each in the new tool's signature workflow
- Replace the floor binder with read-only access to the new tool
- Train the welders on the new floor view
- Retire the Excel files (archive, don't delete)
Plan 2–4 hours per WPS for clean migration. A 30-WPS library migrates in 60–120 hours of engineering time. Spread over a few weeks, it's manageable.
The deepest argument against spreadsheets
Spreadsheets reward individual heroics. The most capable engineer in the shop builds clever templates with embedded formulas. The library works perfectly until that engineer leaves — and then nobody understands the formulas, the references break, and the library degrades. A software-based library doesn't depend on one person knowing where the bodies are buried.