You added automation to your shop floor or engineering workflow. Now leadership wants to know whether it is paying off — and “it feels faster” will not survive a budget review. The answer lives in your automation ROI KPIs (key performance indicators), but only if you captured a clean “before” picture and keep measuring the same way afterward.
This article is the practical companion to our free guide, The Automation KPIs that Make or Break Your ROI. The guide goes deep on each of the eight KPIs. Here, we focus on the groundwork: how to set a baseline, how to track the numbers after go-live, and how to avoid the mistakes that make ROI impossible to prove.
Why a Baseline Comes Before Any ROI Claim
ROI is a comparison. Without a reliable starting point, there is nothing to compare against, and every improvement becomes an opinion. A baseline also protects you from the opposite problem — a new automation project that gets blamed for a dip that was really caused by a seasonal order mix or a staffing change.
A good baseline does three things. It defines each KPI the same way everyone will use it later. It covers enough time to reflect normal ups and downs. And it records the conditions behind the numbers, so you can explain changes instead of guessing at them.
The 8 Automation ROI KPIs at a Glance
These are the eight KPIs covered in the full guide. Use this table as a quick reference for what to measure and where the data usually comes from.
| KPI | How to calculate | Where the data usually lives |
|---|---|---|
| Time Saved | Original Time − Time After Improvements | Time studies, job travelers, ERP labor records |
| Programming Time Per Part | Total Programming Time ÷ Number of Parts Programmed | CAM programmer time logs, job tickets |
| Machine Utilization Rate | (Actual Run Time ÷ Scheduled Production Time) × 100 | Machine monitoring, MES, production schedules |
| Quote-to-Production Lead Time | Production Start Date − Quote Request Date | CRM or quoting system, ERP work orders |
| Engineering Change Cycle | ECO Released or Implemented Date − Change Request Submitted Date | PDM or PLM change workflows |
| Scrap & Rework Rate | (Scrapped or Reworked Units ÷ Total Units Produced) × 100 | Quality logs, nonconformance reports |
| Setup Time Per Job | Total Setup Time ÷ Number of Jobs | Operator logs, machine monitoring (where setup start is captured) |
| On-Time Delivery Rate | (On-Time Units Delivered ÷ Total Units Delivered) × 100 | Shipping records, ERP order history |
Four notes on the formulas:
- Machine utilization — run time ÷ scheduled production time is the availability calculation used in overall equipment effectiveness (OEE). Many shops use “utilization” to mean run time ÷ total calendar (available) time instead, which usually produces a lower number for the same machine. Pick one definition, label it clearly, and keep it consistent before and after go-live.
- Quote-to-production lead time — you can also split this KPI into quote turnaround (quote request to quote sent) and order-to-production (order received to production start) to see where the delay sits.
- Setup time per job — count each job once. If a job runs as multiple operations, such as Op 10 and Op 20, all of its setups roll into that one job rather than counting as separate setups.
- Scrap and rework — count a unit once, even if it is reworked and then scrapped, and consider weighting scrap by cost so one expensive part does not count the same as an inexpensive one.
For why each KPI matters, how automation moves it, pro tips, and what good looks like, download the full guide.
How to Baseline Your Automation KPIs in 5 Steps
Step 1: Start With the Workflow, Not the Metric
Pick the process you plan to automate — CNC programming, quoting, configured product design, change management — and choose the two or three KPIs it affects most. Tracking all eight for every project spreads your effort thin. If you are still deciding where to automate, The Manufacturer’s Workbook to Automation Opportunities walks you through finding friction across engineering, programming, production, data, and quoting.
Step 2: Pull History From Systems You Already Have
Many manufacturers have more data than they realize. Work orders, quality records, shipping history, and PDM workflow timestamps can often give you a starting point without a new tool. Where there is no history, run a manual time study before the change goes live, and run it across enough jobs and shifts to reflect your normal mix rather than a single week.
Step 3: Write Down One Definition for Each KPI
Does “run time” include warm-up cycles? Does “on time” mean the promise date or the original request date, and will you count on-time delivery by units, orders, or order lines? Agree on these definitions with operations, engineering, and finance before you collect a single number, and keep them in one shared document.
Step 4: Capture a Window Long Enough to Be Fair
A single good or bad week is not a baseline. Choose a period that reflects your normal mix of parts, customers, and shifts — often a full quarter or one complete order cycle — and note anything unusual that happened during it.
Step 5: Record the Context Behind the Numbers
Log part mix, order volume, staffing, and any equipment downtime alongside your KPIs. When results change after automation, this context helps you explain what the automation likely changed and what the business changed.
How to Track Automation KPIs After Go-Live
Use the same definitions, the same data sources, and the same reporting cadence you used for the baseline. Changing how you measure in the middle of a project is the fastest way to lose credibility with stakeholders.
Watch trends, not snapshots. Automation often improves over time as teams build templates, refine rules, and trust the new process, so an early reading rarely tells the full story. Review results on a regular rhythm — weekly for shop-floor metrics like setup time and utilization, monthly for cross-functional ones like quote-to-production lead time.
Read KPIs in pairs. Faster programming time per part is good news only if scrap and rework hold steady. Higher machine utilization means more if on-time delivery improves with it. Pairing KPIs keeps one gain from hiding a new problem somewhere else. It also keeps a real improvement from looking like a failure. Utilization can drop even when automation is working — better CAM programming or a more efficient post processor can cut cycle times and raise output while lowering utilization. Track utilization alongside parts produced per scheduled hour, so a drop in utilization is not mistaken for a failure.
Here is a simple worked example with illustrative numbers only. If a machine is scheduled for 40 hours in a week and actually cuts parts for 26 of them, its utilization rate is (26 ÷ 40) × 100 = 65% for that week. Track that same calculation week after week, before and after the change, and the trend becomes your evidence.
To turn KPI gains into ROI, translate them into dollars — for example, hours saved multiplied by your fully loaded labor rate, or avoided scrap valued at its material and labor cost. Then weigh those gains against the full investment, including software, implementation, training, and the internal hours your team spends on the project.
Common Tracking Mistakes That Distort ROI
- Measuring only after go-live — without a “before,” improvements cannot be quantified.
- Moving the goalposts — redefining a KPI mid-project makes old and new numbers incomparable.
- Chasing a single number — one KPI in isolation can look great while another quietly slips.
- Ignoring adoption — automation only pays off when people use it, so low usage often explains flat results.
- Keeping results in a spreadsheet nobody sees — share trends with the people doing the work, not just leadership.
Turning KPI Trends Into Your Next Automation Move
Your KPIs do more than prove ROI on past projects. They point to the next opportunity. Long programming times may signal a case for rules-based CAM such as SOLIDWORKS CAM or CAMWorks. Slow quote-to-production lead times on configured products may point toward design automation with DriveWorks. Long engineering change cycles often trace back to disconnected data, where SOLIDWORKS PDM or the 3DEXPERIENCE platform can help connect workflows. Hawk Ridge Systems sells, implements, and supports these solutions, and our team can help you match the KPI to the right tool.
Before automating anything new, make sure the underlying process is ready. Our guide, Optimize. Automate. Accelerate. Future-Proof Your Manufacturing Operations, covers how to optimize people, processes, and systems first. And for practical advice on getting teams to actually use new automation, read Maximize Adoption of Automation & AI in Daily Workflows: 5 Things Manufacturers Can Do NOW.
Try It With Your Own Data: An AI Prompt for Automation ROI
Want a head start on your own numbers? Copy the prompt below into your AI assistant of choice, fill in the brackets with your data, and leave blank anything you do not have yet. It uses the eight KPIs from this article, and it is built to flag missing data instead of guessing.
Before you paste, remove customer names, part numbers, and pricing, and use your company’s approved AI tool.
You are a manufacturing operations analyst. Help me baseline and track the ROI of an automation project using the 8 automation ROI KPIs below. Use only the data I provide. Do not invent numbers. Do not cite industry benchmarks, averages, or targets unless I provide them. If something is missing, tell me exactly what to collect and where it usually lives. Use a code or calculator tool if available and double-check all arithmetic. ABOUT MY OPERATION - Company type / what we make: [e.g., job shop, CNC machined aluminum parts] - Process we automated (or plan to automate): [e.g., CNC programming, quoting, configured product design, engineering change management] - Automation tool or approach: [e.g., rules-based CAM, design automation, PDM workflows] - Tools we already own (optional): [e.g., SOLIDWORKS, SOLIDWORKS PDM] - Go-live date (or planned date): [date] - Baseline period: [start date to end date] - Anything unusual during the baseline period: [e.g., new customer, staffing change, machine downtime, seasonal spike] MY KPI DEFINITIONS (one agreed definition per KPI) - What counts as "run time": [e.g., spindle cutting only, excludes warm-up] - Utilization is measured against: [scheduled production time (OEE availability) / total calendar (available) time] - What counts as "on time": [promise date or original request date] - On-time delivery is counted by: [units / orders / order lines] - Other definitions we agreed on: [list] MY DATA (formula inputs BEFORE and AFTER go-live; same definitions, same sources) - Before period: [start date to end date] After period: [start date to end date] 1. Time Saved = Original Time - Time After Improvements Task or job measured: [ ] Original Time (hrs per [job / part / week]): [ ] Time After Improvements (hrs per [job / part / week]): [ ] Source: [time studies / job travelers / ERP labor] 2. Programming Time Per Part = Total Programming Time / Number of Parts Programmed Total programming time (hrs): Before [ ] After [ ] Number of parts programmed: Before [ ] After [ ] Source: [CAM logs / job tickets] 3. Machine Utilization Rate = (Actual Run Time / Scheduled Production Time) x 100 Actual run time (hrs): Before [ ] After [ ] Scheduled production time (hrs): Before [ ] After [ ] Parts produced in that time (optional, for parts per scheduled hour): Before [ ] After [ ] Source: [machine monitoring / MES / schedule] 4. Quote-to-Production Lead Time = Production Start Date - Quote Request Date Average and median (days) over [N] orders: Before [ ] After [ ] Source: [CRM or quoting system / ERP] 5. Engineering Change Cycle = ECO Released or Implemented Date - Change Request Submitted Date Average and median (days) over [N] changes: Before [ ] After [ ] Source: [PDM or PLM workflow] 6. Scrap & Rework Rate = (Scrapped or Reworked Units / Total Units Produced) x 100 Scrapped or reworked units (count each unit once): Before [ ] After [ ] Total units produced: Before [ ] After [ ] Source: [quality logs / NCRs] 7. Setup Time Per Job = Total Setup Time / Number of Jobs Total setup time, all setups for those jobs (hrs): Before [ ] After [ ] Number of jobs (count a multi-op job once, e.g., Op 10 + Op 20 = 1 job): Before [ ] After [ ] Source: [operator logs / machine monitoring (where setup start is captured)] 8. On-Time Delivery Rate = (On-Time Units Delivered / Total Units Delivered) x 100 On-time deliveries ([units / orders / order lines]): Before [ ] After [ ] Total deliveries ([units / orders / order lines]): Before [ ] After [ ] Source: [shipping records / ERP] CONTEXT LOG (before and after) - Part mix: [ ] - Order volume: [ ] - Staffing: [ ] - Equipment downtime: [ ] - Adoption (who is actually using the new process, and how often): [ ] WHAT I NEED FROM YOU 1. Pick the 2-3 KPIs that matter most for the process I automated, and explain why. 2. Calculate those KPIs before and after (or all eight, if I provided data for them), showing the math. 3. Read the KPIs in pairs (e.g., programming time with scrap/rework, utilization with on-time delivery, utilization with parts produced per scheduled hour) and flag any gain that may be hiding a new problem. If utilization dropped, check whether shorter cycle times raised output before calling it a failure. 4. Use my context log to suggest what the automation may have changed versus what the business changed. Label these explanations as hypotheses, not findings. 5. Call out tracking mistakes in my data: no true baseline, definitions that changed mid-project, a baseline window that's too short, a single KPI in isolation, or low adoption. 6. Based on the weakest KPIs, suggest where to automate next, considering the tools we already own first. 7. Summarize the results in 5 bullets I can share with leadership, clearly labeling any estimates.
The prompt handles the math. For pro tips and what good looks like for each KPI, download The Automation KPIs that Make or Break Your ROI.
Get the Full Automation ROI KPI Guide
A baseline tells you where you started. The full guide tells you what to do with each number. The Automation KPIs that Make or Break Your ROI breaks down all eight KPIs with why each one matters, how automation helps, a practical pro tip, what good looks like, and how it connects to throughput, labor efficiency, and workflows.
Download the free guide and go from friction to flow.
Frequently Asked Questions
What is the best way to measure automation ROI in manufacturing?
Set a baseline for the KPIs your automation project affects most, then track those same KPIs the same way after go-live. Time saved, programming time per part, machine utilization rate, and scrap and rework rate are common starting points.
How do you calculate automation ROI?
ROI = (financial gains − total costs) ÷ total costs × 100. In plain words, add up what the automation saved or earned in dollars, such as labor hours saved at your fully loaded labor rate and avoided scrap. Subtract everything you spent to get there, including software, implementation, training, and internal time. Then divide the result by those same total costs and multiply by 100 to express it as a percentage.
How long should a KPI baseline period be?
Often a full quarter or one complete order cycle, whichever better reflects your normal mix of parts, orders, and shifts. Avoid basing it on a single unusually busy or slow stretch, and note any unusual events that happened during the window.
How do you calculate machine utilization rate?
Machine Utilization Rate = (Actual Run Time ÷ Scheduled Production Time) × 100. Measured against scheduled production time, this is the availability calculation used in OEE. Many shops instead define utilization as run time ÷ total calendar (available) time, so pick one definition and keep it consistent. Track it by machine, work center, shift, or facility, and watch the trend over weeks rather than relying on a single reading.
How often should you review automation KPIs?
Use a regular rhythm and keep it consistent with your baseline. Weekly reviews work well for shop-floor metrics like setup time and machine utilization, while monthly reviews suit cross-functional metrics like quote-to-production lead time and engineering change cycle.
Which automation KPI matters most?
No single metric tells the whole story. The KPIs are strongest when tracked together, because pairing them shows where friction remains and keeps one improvement from masking a new problem.
What if our KPIs do not improve after we automate?
Check adoption first — automation only pays off when people use it. Then confirm you are measuring with the same definitions and data sources as the baseline, review the context you logged (part mix, order volume, staffing, downtime), and give the new process time, since automation often improves as teams build templates and refine rules.
What if we never measured anything before we automated?
Start now. Use whatever historical records you have — work orders, quality logs, shipping history — to reconstruct an approximate baseline, label it as an estimate, and measure consistently from here forward.
How can a manufacturer find its biggest automation opportunities?
Benchmark your current KPIs, then look for the processes with the longest programming times, lowest utilization, or highest scrap. The Manufacturer’s Workbook to Automation Opportunities is a practical place to start. For an outside perspective, a Hawk Ridge Systems Business Assessment reviews your product development processes, pinpoints inefficiencies, and delivers a plan to address them.
Contact Us
Not sure which KPIs to baseline first, or where automation will make the biggest difference in your operation? Contact Us to talk with a Hawk Ridge Systems expert about your workflows, your data, and a practical plan to measure the results — or ask about a Business Assessment to uncover inefficiencies and get a plan to close them.
