Metric-driven decision making for manufacturing production
Question - How often does your production team see and respond to their performance metrics?
📊 Real-time on shop floor?
📱 Daily in team huddles?
📑 Weekly reports?
📅 Monthly reviews?
Without frequent reviews of performance and visibility into production data, production teams are destined to just do their own thing, and this might (* definitely) not be what you are looking for as a manager.
If you’ve read ‘How to Win Friends & Influence People’ by Dale Carnegie you may remember the story of the industrialist Charles Schwab visiting the Bethlehem Steel Mill because the mill was continually not hitting quota. To cut a long story short, at the end of the day Schwab visited he wrote the number of batches of steel in chalk on the floor (I think the number was 6).
When the next shift came in, they asked what the number was, and got the answer that ‘the big boss’ had asked and written it down.
The next morning Schwab came back to see the 6 replaced with a 7. The night shift had upped the ante. The day shift then came in, and competitive instincts raised, put in the effort to make it to 10. Nearly twice the amount of the day before. (I think this was still behind quota, but still a massive improvement). After a while, these improvements continued to make the plant the most productive.
With overall gross margin improvement as a goal, where should you start?
Firstly - tracking at least weekly. Weekly is often a good cadence for management cycles. Weekly gives you pattern visibility as well as time for analysis without taking too much time from the team, but if issues are only raised weekly, there can be a delayed response compared to daily, or forgetfulness of what actually happened. So at least track weekly, but don’t only act weekly.
Secondly, which metrics should you start with? Your production team can’t control some parts of COGS e.g. raw materials or procurement, but there are a few it can control.
A good five to start with are First Pass Yield, Machine Downtime, WIP Aging, Changeover Time and Energy usage per unit. These directly can impact reducing COGS, improving cashflow and customer experience when managed in the right direction.
First Pass Yield (FPY).
FPY = (Total Units - Defect units - Reworked units) / Total Units Started) × 100%
First Pass Yield is an essential quality metric and defined by ISO 22400-2. Higher is better, as long as you are doing the checks to identify defects or rework required. (Don’t allow it to be gamed it in a way that you don’t check quality). It is a better metric than simply total units throughput because of its inspected quality element.
Increasing FPY (i.e nearer to 100%) directly impacts COGS in terms of direct material, labour and production overhead time. In some situations it may also improve customer satisfaction by shortening lead times, reducing issues with units that get to customers and so on.
Impact of Low FPY = Base COGS × (1 + (1-FPY) × Rework Factor)
Where Rework Factor includes:
Material waste %
Additional labor %
Extra overhead %
Quality management costs
There are a number of sources needed to get this data, depending on the complexity of your management process, including your Quality Management System, manual checks, SPC (statistical process control) systems and machine PLC systems, meaning a way to pull this data together automatically is crucial.
For each and every metric, you need to clarify in detail what each of the constituent parts of the metrics means, what is included, and what is not, before starting measurement (and then refine every so often). For example, define exactly what you call a defect - i.e. units that do not meet spec and cannot be reworked - and what exactly that means in your organisation. Define exactly what you mean by ‘started’ and when it is started . Document each time point in the process and what it means. Document what steps are included.
Machine Availability
Availability = (Actual Run Time (= Planned production time - Unplanned Stops)) / Planned production time * 100%
Planned production time = Total time (=24*7) - Planned stops
Machine availability is a core component of OEE (Overall Equipment Effectiveness), defined by ISO 22400-2. Higher is better. It's better than simple utilisation because it focuses on stoppages you can actually improve. Make sure that you categorise stoppages correctly and are agreed on the reasons, otherwise you won’t. Balance with maintenance schedules so that unplanned stops are reduced.
Improving availability directly impacts COGS through better absorption of fixed costs and reduced overtime. It can improve customer satisfaction through more reliable delivery times and consistent quality.
Impact on Cost = Machine Hour Rate × Unplanned Stop Hours × Loading Factor where Loading Factor includes:
Overtime premium
Lost capacity cost
Catch-up inefficiency
Additional setup/startup waste
Data sources needed include OEE systems, SCADA, PLC feeds, operator inputs, and CMMS. Key is accurate categorisation of:
Planned stops (maintenance, breaks)
Unplanned stops (breakdowns, shortages)
Running time
Also, if you’re running the Andon cord principle, be careful not to penalise people for doing the right thing - so categorise stops appropriately.
WIP Aging
WIP Days = Current WIP Value / Average Daily COGS
or
MCT (Manufacturing Cycle Time) = Current Inventory / Average Daily Production
WIP aging is a key cash flow metric defined in the SCOR model. Lower is better - every day of WIP is cash trapped in your process. It's better than total WIP value alone because it shows how fast work moves through your system.
Reducing WIP days directly impacts cash flow and COGS through reduced handling, space usage, and quality risk. It often improves customer satisfaction through shorter lead times and fresher product.
Impact on Working Capital = WIP Days × Daily Production Cost × Carrying Cost Factor
where Carrying Cost Factor includes:
Storage cost %
Capital cost %
Handling cost %
Obsolescence risk %
Data sources include ERP, MES, WMS and production scheduling systems, requiring consistent timestamp tracking at each process step.
Changeover Time
Changeover Time = Time from Last Good Piece of Previous Run to First Good Piece of Next Run
Changeover time is a critical flexibility metric under SMED methodology. Lower is better, but don't sacrifice startup quality. It's better than simple 'setup time' because it includes stabilization time to good product.
Reducing changeover time impacts COGS through increased capacity and reduced batch sizes. It improves customer satisfaction by enabling more flexible production scheduling.
Impact on Cost = (Changeover Hours × Machine Rate) + (Startup Scrap × Unit Cost)
where:
* Machine Rate includes labor and overhead
* Unit Cost includes materials and value-add
Data needed from MES, OEE systems, quality systems and operator inputs, requiring clear definition of 'good piece' criteria.
Energy Usage per Unit
Energy Performance Indicator (EnPI) = Total Energy Consumed / Units of Production
Energy per unit is an ISO 50006 efficiency metric and should be managed from a COGS as well as regulatory perspective. As well as addressing COGS through energy cost reduction, it is often needed for environmental reporting such as GRI 302) Lower is better but watch for impact on quality and throughput. It's better than total energy cost because it accounts for production volume changes.
Reducing energy per unit directly impacts COGS through utility costs and often indicates better equipment health. Can improve sustainability metrics for customers and investors, who increasingly have sustainability targets, so improving energy efficiency can make you more investable.
Impact on Cost = Units × (Current EnPI - Target EnPI) × Energy Rate
where Energy Rate includes:
Base utility cost
Peak demand charges
Environmental levies
Carbon costs
Data sources include energy monitoring systems, smart meters, PLC/SCADA systems and production counts, requiring time-synchronised data collection.
Summary
Being a data driven production shop requires at least weekly tracking of the key metrics that matter. A good 5 to start with are those presented here.
First Pass Yield, Machine Downtime, WIP Aging, Changeover Time and Energy usage per unit are a great starting point. Don’t forget to be really specific about what you mean for each with all the parts that build into the formulae. Which exact time point, which exact status, which observation and so on.
In order to achieve this, integrating your systems’ data automatically is crucial to being able to manage your COGS, cashflow and improve customer experience and provide external reporting.
How Calon helps
Calon gets the metrics at the heart of your business in front of the people that matter, at the right time. We call this the Operating System for growing B2B / multichannel companies.
We can take data from any system, combine it with the rest of your data and automatically produce the insights you need. We also help you get all your definitions right, improve your data quality and get your teams working together with their data to drive your business.
Visit us at https://www.calonanalytics.com, follow us on linkedin at https://www.linkedin.com/company/calonanalytics/ or drop me a mail at nathan@calonanalytics.com
