You have a process that's done manually. It costs money. Everyone knows this. But you don't know how much.
So when someone proposes automation, you don't know if it's worth it.
Here's how to calculate the true cost of a manual process.
Next Steps
Once you understand the true cost, evaluate automation opportunities and consider how operations data can guide your choices.
The Total Cost Calculation
Manual processes have three cost components:
Direct cost: labor to perform the work Indirect cost: systems, infrastructure, space needed for the work Opportunity cost: what else could that labor do?
Most companies only calculate the direct cost.
Direct Cost
This is straightforward: how many person-hours does the process consume?
Example: invoice processing
Current state:
- 50 invoices per week arrive (email, mail, EDI)
- One person processes them
- Takes 10 minutes per invoice
- 50 × 10 minutes = 500 minutes per week = 8.3 hours per week
- One person could almost handle it, but they also do other things
- Actually, it's 1.2 FTE (one person full-time plus 20% of another person's time)
Annual cost:
- 1.2 FTE × $80K (salary + overhead) = $96K per year
That's the direct cost everyone knows about.
Indirect Cost
But there are costs hidden in infrastructure:
Error handling: Invoices have errors. Wrong quantities, wrong prices, missing information. Someone spends 5% of time fixing errors.
Time: 8.3 hours/week × 5% = 0.4 hours/week = 20 hours/year
Cost: 20 hours × $80/hour = $1,600/year
Escalation: Invoices from new vendors or unusual orders need approval. Manager spends 2 hours per week reviewing and approving.
Cost: 2 hours/week × $80/hour = $8,320/year
Tracking: Someone needs to track which invoices have been processed, which are pending. Informal tracking in email, missing some.
Time: 3 hours/week = 150 hours/year
Cost: $12,000/year
Rework: Invoices that were processed wrong and discovered later. 2% of invoices need reprocessing.
Time: 50 invoices/week × 2% × 20 minutes average rework = 20 minutes/week = 17 hours/year
Cost: $1,360/year
System/infrastructure: The email systems, file storage, spreadsheets. Not huge, but some cost.
Cost: $2,000/year
Total indirect cost: $25,280/year
Opportunity Cost
The people doing the manual work could be doing something more valuable.
If the invoice processor wasn't spending 8.3 hours per week on invoice processing, they could be:
- Working on vendor relationship management (talking to vendors, resolving issues early, negotiating better terms)
- Analyzing spending trends
- Improving processes
- Other value-add work
What's that work worth?
If they move from "process invoices" to "manage vendor relationships," they might generate $30K-$50K annually in better terms, faster resolution, fewer disputes.
But you don't capture that benefit because you don't see it as an alternative use of time.
Still, it's a real opportunity cost.
Let's be conservative: $15K per year in unrealized value from doing better work.
Total Annual Cost
Direct cost: $96K Indirect cost: $25.3K Opportunity cost: $15K
Total cost of manual invoice processing: $136K per year
But the person's salary is only $96K. The other $40K in costs are hidden.
The Automation Decision
Now you can evaluate: is automation worth it?
Automation options:
Option 1: Partial automation (OCR + validation)
- Cost: $5K setup
- Annual maintenance: $2K
- Result: OCR extracts data from invoices, system validates, human reviews
- Labor savings: 7 of 8.3 hours per week (most invoices auto-processed, humans spot-check)
- New labor requirement: 0.15 FTE (human review only)
- Labor cost saved: 1.05 FTE × $80K = $84K/year
- Remaining costs: $2K/year + $12.3K (error handling, escalation, tracking for remaining manual work)
Net benefit: $84K labor saved - $2K maintenance - $12.3K remaining costs = $69.7K/year
ROI: $5K investment, $69.7K annual benefit, payback in 26 days
Option 2: Full automation (intelligent invoice processing)
- Cost: $30K setup + $500/month SaaS = $36K first year
- Result: invoice arrives → system extracts, validates, matches to PO, posts to GL, sends payment → minimal human intervention
- Labor savings: 1.2 FTE = $96K/year
- Remaining costs: $5K (exception handling) + $1K (vendor management)
- New labor: 0.05 FTE for exception handling
Net benefit: $96K labor - $36K automation cost - $6K remaining = $54K/year
ROI: $36K investment, $54K annual benefit, payback in 8 months
The Full Calculation Template
For any manual process:
- Document the process: who does it, how long does it take?
- Calculate direct cost: person-hours × hourly cost
- Identify indirect costs:
- Error handling (time to fix mistakes)
- Escalation (time for approval/review)
- Tracking/reporting (time to track progress)
- Rework (time to fix discovered errors)
- Infrastructure (systems, space, tools)
- Estimate opportunity cost: what better work could this person do?
- Total cost: direct + indirect + opportunity
- Evaluate automation:
- Cost of automation (setup + annual)
- New labor requirement (exception handling)
- Compare: total cost of manual vs. total cost of automated
- Calculate ROI: (savings - automation cost) / automation cost
Real Examples
Example 1: Expense Reports
Manual process:
- Employees submit reports by email with receipts as attachments
- Manager reviews (5 minutes per report)
- Finance person processes in accounting system (3 minutes per report)
- Finance manager audits and approves (2 minutes per report)
- Accounting posts and sends to employee
Volume: 100 reports per month
Direct cost:
- Employees: 100 × 20 minutes/month prep = 33 hours/month = $3,960/year
- Manager review: 100 × 5 minutes = 8 hours/month = $960/year
- Finance processing: 100 × 3 minutes = 5 hours/month = $600/year
- Manager approval: 100 × 2 minutes = 3 hours/month = $360/year
- Total: $5,880/year
Indirect costs:
- Errors (resubmissions): $1,500/year
- Tracking overdue reports: $800/year
- Rework due to missing receipts: $2,000/year
- Total indirect: $4,300/year
Total manual cost: $10,180/year
Automation (expense management software):
- Cost: $3K setup + $200/month = $5,400 first year
- Result: employees submit via app, receipts auto-captured, OCR extracts data, manager approves with one click, posts automatically
- New labor: 1 hour/month for exceptions = $120/year
- Error reduction: $500/year
Net benefit: $10,180 - $5,400 - $500 - $120 = $4,160/year
ROI: $5,400 investment, $4,160 annual benefit, payback in 15 months
This passes the ROI test (positive ROI, reasonable payback).
Example 2: Customer Onboarding
Manual process:
- Sales sends customer info
- Operations person sets up account (30 min)
- Operations sends welcome email (10 min)
- Operations invites to kickoff meeting (10 min)
- Operations notifies IT
- IT sets up access (1 hour)
Volume: 20 customers per month
Direct cost:
- Operations: 20 × 50 minutes = 16.7 hours/month = $2,000/year
- IT: 20 × 60 minutes = 20 hours/month = $2,400/year
- Total: $4,400/year
Indirect costs:
- Errors (wrong access, missing info): $3,000/year (rework time + delays)
- Tracking delays: $1,000/year
- Total indirect: $4,000/year
Total manual cost: $8,400/year
Automation (workflow automation tool):
- Cost: $8K setup + $300/month = $11,600 first year
- Result: customer info → system sets up account → sends emails → invites to meeting → notifies IT → IT creates access (all automated)
- New labor: 1 hour/month for edge cases = $120/year
- Error reduction: $2,500/year
Net benefit: $8,400 - $11,600 - $120 + $2,500 = -$820/year
ROI: Negative in year one, breaks even in year two if the $11,600 first-year cost is just setup
This is borderline. It makes sense if:
- You expect volume to grow to 50+ customers/month (then savings are $20K/year)
- Or if the error reduction and customer satisfaction improvement are worth more than the $820 cost
When Manual Makes Sense
Sometimes manual processes are actually cheaper or better:
- Very low volume (<5 per month) where automation overhead is too high
- Highly variable or unpredictable work where every case is different
- Judgment-based decisions where human discretion is essential
- Situations where you don't have good data to automate
In these cases, don't automate. Just improve the manual process.
When Automation Makes Sense
Automate when:
- ROI is positive within 12 months
- Process is rule-based and predictable
- Volume is high enough to justify the cost
- Process is repeated frequently
- Or the cost of errors is very high (even low volume might make sense)
Use the calculation to decide, not gut feel.
Most manual processes that feel painful actually cost 2-3x what people think.
Calculate the real cost, and automation usually makes financial sense.
Calculate Your Real Costs
If you have manual processes that feel expensive, we can help you quantify the true cost and evaluate automation options. Book a discovery call to run the numbers on a specific process and understand whether automation makes sense for your situation.
