The Real ROI of PCB Assembly Automation: A 12-Month Payback Model You Can Run in 10 Minutes
Diagnosis
Automation business cases fail in two opposite directions, and both are common. The optimistic case counts only direct labour saved, assumes the machine runs at rated speed on day one, and ignores feeders, grippers, installation, training and spare parts — so the project is approved and then misses its payback target by a year. The pessimistic case counts the full capital cost but only counts labour at base wage without overtime, shift premiums, recruitment, training and attrition — so a project that genuinely pays back in twelve months gets rejected. Neither error is a spreadsheet error. Both are scoping errors.
Step 1 — Count the fully loaded labour cost, not the wage
For every operator the automation displaces or redeploys, add employer contributions, shift premium, overtime, recruitment cost amortised over expected tenure, training time and attrition replacement. In many markets the fully loaded figure runs 1.4 to 1.9 times base wage. Using base wage alone is the single most common cause of rejecting a good project.
Step 2 — Count the full capital cost, not the machine price
Add feeders and grippers, tooling, installation and commissioning, operator and technician training, spare parts kit, freight, duties and local registration. A realistic figure is machine price plus 15 to 25%. Using machine price alone is the single most common cause of approving a bad project.
Step 3 — Use achievable throughput, not rated throughput
Model at 60 to 70% of rated CPH for mixed production in the first six months, rising to 75 to 85% as programmes mature. If your business case only works at 100% of rated speed, it is not a business case.
Step 4 — Quantify yield and rework separately from labour
Automated insertion typically reduces polarity and bent-lead defects substantially compared with manual work. Estimate the reduction in defects per million, multiply by your rework cost per defect inclusive of test time and scrap. On complex boards this line item is frequently 20 to 30% of total benefit and is almost always omitted.
Step 5 — Add the option value of capacity
What is the margin on the incremental volume you could accept if the line were not labour-limited? If you are currently turning away orders or capping volume because you cannot hire, that has a value. It does not belong in a conservative base case, but it belongs in the sensitivity analysis that gets the project approved.
Step 6 — Run the sensitivity, not just the point estimate
Vary three inputs: throughput at 60% and 85%, labour cost at plus and minus 20%, and capital cost at plus 15%. If the project pays back inside 24 months in the worst case, it is robust. If it only works in the best case, say so out loud before signing.
Our vehicle
We run this model with customers before quoting, not after. Send us your BOM, board drawing and volumes, and we will return a completed model with our numbers and assumptions exposed so your finance team can challenge every line. That is also why our quotations itemise feeders, grippers, training and spare parts separately — a quote that hides those lines is a quote you cannot audit, and a business case you cannot audit is one you cannot defend internally.
The compounding loop
Keep the model alive after installation. Log actual throughput, actual labour redeployment and actual defect rates monthly against the assumptions you approved. Two things happen when factories do this: the next business case gets approved faster because prior estimates were honest, and the actual constraint on the line becomes visible, which is where the next automation project should start.
Download the THT staffing audit calculator, fill in one board part number, and send it back to us. We will complete the payback model with your numbers and return it with every assumption visible.
What payback period should I target for PCB assembly automation?
Most EMS customers target 12 to 24 months. Insertion and soldering equipment with high labour content frequently lands in the 9 to 18 month range. Anything beyond 36 months is usually a sign the configuration is wrong rather than the technology being unsuitable.
Should I use base wage or fully loaded labour cost?
Fully loaded cost for the business case, because that is the cash the company actually stops spending. Fully loaded figures typically run 1.4 to 1.9 times base wage once employer contributions, shift premiums, overtime, recruitment and training are included.
How much should I budget above the machine price?
Plan on 15 to 25% above machine price to cover feeders, grippers, tooling, installation, commissioning, training, spare parts and freight. Ask your supplier to itemise these separately so you can audit them.
Is it better to buy new or refurbished equipment?
Refurbished equipment can make sense for standard processes like wave soldering and reflow, provided you verify support and spare parts availability for the specific platform. For insertion and odd-form work, where feeders and grippers are application-specific, new equipment with a validated trial usually carries lower project risk.
