Tangle Research
How AI Quoting Changes Win Rates in Custom Metal Fabrication
A $5M-revenue, 25-FTE custom metal-fab shop sending 60 quotes a month at 25% win rate writes $4.5M of won-quote revenue a year. AI-native quoting that lifts capacity by 20% and win rate by 3 points adds $1.55M of revenue and $387,000 of margin in year one.
1. The bottleneck is the estimator, not the machine
Walk through any custom metal-fab shop and the operations leader will point at the press brake, the laser, the welding cell. Somewhere on the floor. Ask where revenue actually stalls and the honest answer is upstream. At the estimator's desk.
Quoting custom work is hard. Every job is different. Geometry, material, finish, tolerances, volume, lead time, the customer's quirks — all of them affect cost. The knowledge sits in heads and spreadsheets. The estimating team is rarely large. Quote volume scales with sales activity. Estimating capacity does not. Quotes go out late, or do not go out at all.
RFQ response time is one of the most consistent predictors of win rate in custom fabrication. Customers send the same RFQ to three to five shops. The first credible quote often wins. Slow quoting compounds twice. Fewer quotes get sent at all. And the ones that do get sent arrive after the customer has already started leaning toward someone else.
2. What AI-native quoting actually does differently
Traditional quoting asks the estimator to read the print, identify operations, look up rates, apply markups, write the quote, send it. AI-native quoting changes the economics of each step.
Geometry parses from CAD automatically. Operation sequences come from feature recognition or from historical jobs that match the part. Material costs pull from live supplier data and updated tariff or surcharge tables. Labor estimates come from actual machine cycle times in the shop's own historical data, not from rules of thumb. The estimator's role shifts. From building the quote from scratch, to checking the AI's draft and applying commercial judgment.
Published case data is consistent. Time per quote falls 30 to 50% with AI-native quoting deployed properly. Tangle's own number is a 90% reduction in quoting time. Accuracy improves at the same time, because the numbers come from real data rather than estimator memory.
3. The math: capacity uplift and win-rate uplift
Two effects drive the dollar impact.
Capacity uplift. The same estimating team sends more quotes. If time per quote drops 30 to 50%, conservative capacity uplift sits at +10 to +20%. Most of the time gain gets absorbed by quality checking rather than pure quote volume, in honest accounting.
Win-rate uplift. Faster response and tighter pricing convert at a higher rate. A 1 to 3 percentage-point shift is the defensible range. Modest in absolute terms. Large in dollar terms.
Annual won-quote revenue = quotes per month × 12 × win rate × average deal size. The improved version multiplies quotes by (1 + capacity uplift) and adds the win-rate uplift to the conversion percentage. Margin contribution is the revenue delta times gross margin. The model uses 25% — the FMA average for custom job shops.
4. Worked example: $5M, 25-FTE shop
Baseline: 60 quotes per month × 12 × 25% win × $25,000 average deal = $4.50M of annual won-quote revenue.
| Scenario | Capacity uplift | Win-rate uplift | Won-quote revenue | Δ revenue | Δ gross profit |
|---|---|---|---|---|---|
| Baseline | — | — | $4,500,000 | — | — |
| Conservative | +10% | +1 pt | $5,148,000 | +$648,000 | +$162,000 |
| Realistic | +20% | +3 pts | $6,048,000 | +$1,548,000 | +$387,000 |
5. What the data says
Published outcomes from job-shop ERP deployments cite quoting time reductions of 30 to 50% as routine. Tangle's own deployments come in at 90%. Win-rate uplift is less commonly published as a clean number, because shops rarely measure it before deployment. But RFQ-response-speed analysis from industry trade publications consistently lists "first credible quote" as a primary buying factor for custom fabrication customers.
The conservative case here — +10% capacity and +1 percentage point on win rate — already produces a $162,000 annual margin gain on a $5M shop. The realistic case at +20% and +3 points produces $387,000. Both sit inside the published range for shops that deploy an AI-native ERP with discipline and data hygiene.
Quoting is rarely the bottleneck shop owners point at. It is almost always the bottleneck the numbers point at.
Run the model with your own numbers
Three to five minutes. Five inputs. Same framework, applied to your shop.
Open the ROI calculator· Frequently asked questions
How much does AI quoting reduce time per quote in a metal-fab shop?
Published case data from job-shop ERP deployments shows quoting time per quote falling 30 to 50%. Tangle's own number is a 90% reduction. Most of that time saving gets absorbed into higher-quality quote review, so the practical capacity uplift sits in the +10 to +20% range.
Does AI quoting actually improve win rate, or just speed?
Both, for different reasons. Speed improves win rate because RFQ response time is a primary buying factor in custom fabrication. Accuracy improves win rate because more disciplined cost estimating produces sharper pricing — too-high quotes lose deals, too-low quotes lose margin. A +1 to +3 percentage-point uplift is the defensible range.
What gross margin should I assume in the dollar calculation?
FMA-reported average gross margin for custom job shops sits around 25%. Higher-margin shops (specialty alloys, tight-tolerance precision work) run 30 to 35%. Commodity laser-cut work runs lower. Use the shop's own number if available.
Does this work for shops with low quote volume?
Yes. Capacity uplift matters less for shops that are already keeping up, but accuracy uplift still affects win rate. Shops with low quote volume often gain more from the accuracy effect than the capacity effect.
· Sources
- FMA Financial Ratios & Operational Benchmarking Survey — Industry gross margin and labor benchmarks for custom job shops.
- The Fabricator — Estimating basics and quoting jobs in custom metal fabrication — Quoting workflow context and accuracy drivers.
- The Fabricator — Metal fabrication estimating strategies that drive profitable job shop growth — Win-rate framing and RFQ-response context.
- Tangle deployment data — 90% reduction in quoting time across customer deployments.