The Tool 80% of Steel Shops Actually Use
When people talk about "steel takeoff software," they usually mean PlanSwift, Bluebeam, or one of the AI tools entering the market. But the honest reality is that most small-to-mid steel fabrication shops do their takeoffs in Excel.
The workflow looks like this: open the PDF in a viewer, scroll through each page, write down every steel member you find, and type it into a spreadsheet. Maybe you have a template with weight-per-foot formulas. Maybe you have a tab for each project. Maybe it's the same file your predecessor built 15 years ago.
It works. It's been working for decades. The question isn't whether Excel can do steel takeoffs — it's whether it should, given what's available now.
The Excel Takeoff Workflow
Here's what a typical Excel-based steel takeoff actually involves:
- Print or open the PDF — many estimators still print sheets and mark them up with highlighters
- Page-by-page scan — scroll through every structural page, visually identify every steel callout
- Manual entry — type each member designation into the spreadsheet: W12X26, HSS6X6X1/4, C10X15.3
- Weight lookup — reference the AISC shape database (book, PDF, or a lookup tab in the same spreadsheet)
- Quantity aggregation — count occurrences, sum lengths, calculate total weight
- Formula maintenance — keep the spreadsheet's formulas, lookups, and formatting working correctly
- Cross-check — go back through the drawings to make sure you didn't miss anything
For a typical 20-page structural package, this takes a skilled estimator 3-6 hours. For large projects (50+ pages), it can take days.
The Steelflo Workflow
- Upload the structural PDF
- Set the drawing scale
- AI detects every steel member — W-shapes, HSS, channels, angles, pipes, plates, lintels
- Review detections on the source page with bounding box overlay
- Confirm, reject, or adjust
- Export BOM, CSV (opens directly in Excel), order sheet with cut optimization
For the same 20-page package, detection runs in minutes. Review and verification adds 15-30 minutes. Total: under an hour.
Side-by-Side Comparison
| | Steelflo | Excel Takeoff | |---|---|---| | Detection | AI-powered automatic | Estimator's eyes + highlighter | | Member identification | Automatic classification | Manual entry, one at a time | | Weight data | Built-in (AISC, EN, BS, AS/NZS) | External reference or lookup formulas | | Source traceability | Every detection linked to page + bounding box | Depends on estimator's notes | | Error risk | AI may miss ambiguous callouts | Missed members, typos, miscounts | | Time (20-page set) | Under 1 hour | 3-6 hours | | Time (50+ pages) | 1-2 hours | 1-3 days | | International standards | Auto-detected (4 standards) | Estimator must know the notation | | Cost | $399–$1,499/mo | Free (you already have Excel) | | Learning curve | Low (6-step wizard) | None (you already know it) | | Flexibility | Structured workflow | Total flexibility | | Nesting/cut optimization | Built-in | Build it yourself or use another tool |
Why Excel Still Works (And When It's Fine)
Let's be real about when Excel is the right call:
Small projects with few members. A 3-page structural set with 15 steel members? Excel is faster than uploading to any software. The overhead of any tool exceeds the time saved.
One-off estimates. If you do 2-3 steel bids per month, the monthly cost of specialized software might not pencil out. Excel is free and you already know it.
Custom calculations. Excel gives you infinite flexibility. Custom waste factors, shop-specific markup formulas, pricing models that no off-the-shelf tool supports — Excel handles all of it.
Legacy workflows. If your shop has a 15-year-old estimating spreadsheet that's been refined through hundreds of bids, there's institutional knowledge embedded in those formulas. That's valuable and hard to replicate — though if those files aren't organized and backed up, you're one hardware failure from losing all of it.
Where Excel Breaks Down
Scale. A 20-page structural package is tedious. A 50-page package is brutal. The manual identification step doesn't scale — every additional page means more scrolling, more counting, more data entry. The most common takeoff mistakes — missed members, double-counts, transposed designations — get more frequent as the page count grows.
Speed under pressure. When the GC calls at 2 PM on Friday and needs a number by Monday, the Excel workflow becomes the bottleneck. You can't accelerate manual page scanning. Steelflo runs the detection in minutes regardless of page count — fast turnaround is the whole point.
Auditability. In Excel, your audit trail is whatever notes you chose to write. If someone asks "where did this W14X22 count come from?" you're flipping through marked-up prints or relying on memory. In Steelflo, every detection links to a specific page and bounding box on the source drawing.
International work. If a drawing uses EN notation (HEA200, IPE300) or AS/NZS notation (310UB40.4), your Excel lookup table probably doesn't have those sections. Steelflo's databases cover all four major international standards.
Estimator dependency. The Excel workflow lives in the estimator's head — and usually on one computer. If your estimator leaves, gets sick, or retires, the institutional knowledge goes with them. AI-assisted takeoff creates a repeatable, documented process that doesn't depend on one person's experience.
The Real Cost of Excel
Excel is free. But the takeoff process isn't.
An experienced steel estimator billing at $50-70/hour spending 5 hours on a takeoff costs $250-350 in labor. At 15 bids per month, that's $3,750-5,250 in estimating labor — for the takeoff step alone, before pricing, bidding, and negotiation.
If Steelflo cuts that to 1 hour per takeoff, you save 4 hours × $60 × 15 bids = $3,600/month in labor against a $399-599/month subscription.
The math only breaks down at very low bid volumes (under 3-4 per month) or very small projects (under 5 pages). For everything else, the labor savings exceed the software cost by a wide margin.
The Hybrid Approach
Many steel shops don't make a hard switch. They use Steelflo for detection and quantity extraction, then export to CSV and open it in their existing Excel template for pricing, markup, and custom calculations.
This gives you the speed of AI detection with the flexibility of your established spreadsheet workflow. The export is designed for exactly this — structured data that drops into whatever downstream process you already have.
Is Excel Good Enough for Steel Takeoffs?
For small projects and low bid volumes, yes. For larger packages (20+ pages) and shops doing 10+ bids per month, the manual identification step becomes the bottleneck that limits how many bids you can pursue.
How Much Time Does Steelflo Save Over Excel?
On a typical 20-page structural package, Steelflo reduces the takeoff from 3-6 hours to under 1 hour. The time savings increase with drawing set size — 50+ page packages show the biggest difference.
Can I Still Use Excel with Steelflo?
Yes. Steelflo exports to CSV, which opens directly in Excel. Most shops use Steelflo for detection and quantity extraction, then import the data into their existing Excel templates for pricing and custom calculations.
What If My Excel Template Has Custom Formulas?
Keep using it. Steelflo handles the detection and quantification step — the part that takes the most time. Export the results to CSV, import into your template, and your custom formulas, waste factors, and markup calculations still apply.
Is Steelflo Worth It for a Small Shop?
Depends on volume. If you do fewer than 3-4 steel bids per month on small projects, Excel is probably fine. Above that threshold, the labor savings typically exceed the subscription cost. The free trial lets you test on a real drawing before committing.
Try Steelflo on a Bid You Already Estimated
The most convincing comparison is your own. Upload a structural PDF you've already taken off in Excel. Compare the member counts, check the weights, and see how long it takes.
Start free at steelfloai.com — one AI takeoff, no credit card.