Case Study · Packaged Food Manufacturer
From siloed spreadsheets to 37.7% less loss and waste.
How we aligned Finance, Operations, Sales, and Warehouse around one definition of loss and waste. The reduction is measured like for like, with commodity price movement and a one-time acquisition write-off stripped out of the comparison.
37.7%
Normalized YoY reduction
4 of 4
Loss accounts improved
8 wks
To one reconciled number
The Challenge
"What is our actual loss and waste?" depended on who you asked
Finance
The general ledger
One piece of it
Operations
Production reports
One piece of it
Sales
Customer credits
One piece of it
Warehouse
Inventory variance
One piece of it
When we first met with the leadership team, they had a problem they could not quite name. The CFO could tell you the P&L impact of loss and waste. Operations could tell you production-floor inefficiencies. The warehouse team could tell you inventory variance. Sales could tell you customer credits. Each of them was right about their piece. Nobody owned the whole. There was no shared definition of loss and waste across the company, no agreed scope, and no way to hold the total against a budget.
Each function had built its own tooling to answer its own questions, and each report was defensible on its own terms. Sales tracked credits. Operations tracked scrap. Neither had a reason to count the other's. Added together they did not make a company number, so a meeting about reducing loss became a meeting about what counted.
The Solution
One definition, calculated from actual data
We started with a question: what is loss and waste as you see it? We brought Finance, Operations, Warehouse, and Sales together and asked each group to define it. Then we asked: if we add all these up, where do they reconcile? They didn't, and the gap between the highest and lowest view was wide enough to change decisions.
Over a month of weekly sessions, we arrived at a clean separation, and assigned owners: Operations Manager, Warehouse Manager, Sales Director, Procurement Director, CFO. Using actual data from the GL, operations reports, and sales records, we calculated the truth. For the first time, Finance, Operations, and Sales were looking at the same number.
Controllable loss
Loss where a specific owner could take action: production scrap, inventory damage, expired goods, rework, process overages.
Non-controllable loss
Loss that could be measured but where the owner operated within constraints set by others: customer credits, standard-cost variance, regulatory disposal.
Process Redesign
Beyond metrics: fixing the processes that feed them
Standardized reason codes
Warehouse staff used to enter free-form notes: "Damaged this item." "Short-coded." "Missing." We built a required dropdown with standardized reason codes: cycle count, damage in-house, damage vendor, discontinued, donations expired.
What was invisible became sortable and actionable. Loss could finally be ranked by cause, and each category got an owner responsible for reducing it.
A/B/C/D cycle counting
We implemented a cycle-counting program in their NetSuite ERP using A/B/C/D inventory priority. A items counted daily, B weekly, C monthly, D quarterly.
Active monitoring replaced the reactive inventory close. Ongoing trust in system numbers, and no more month-end arguments about mismatches.
The Results
Year over year
-66.0%
Customer credits
-43.7%
Inventory adjustments
-16.4%
Donations
-6.7%
Inventory damages
Every loss account improved, not just one. Together they came to a 37.7% reduction year over year. Within four reviews (eight weeks), the metrics stabilized: Finance, Operations, and Sales pulled the same loss and waste number each week. Within three months, all departments tracked against specific improvement plans.
How We Measured It
We could have claimed 61%. Here is why we publish 37.7%.
Taken at face value, the general ledger shows loss and waste down more than 60% year over year. Two things in that number were not earned on the plant floor, so we took them out before publishing anything.
Every loss account, as posted to the GL.
Milk variance tracks the commodity market, not operations. The write-off was inventory booked at an acquired facility that did not physically exist. Neither was earned on the plant floor, so both come out.
Why this matters to you
A number you cannot defend is worse than no number. The figure we publish is the one that survives a like-for-like comparison: same months, same locations, same accounts, both years.
Every figure on this page is rebuilt from the client's own general ledger and reconciles line by line. The facility that closed between the two years is excluded from both, so nothing is credited to a plant that stopped reporting.
That is the same standard we apply to the reporting we build. If a metric cannot be traced back to a source, it does not ship.
Same Client, Second Number
99% or 93%? Both were true.
On-time-in-full at the same manufacturer runs about 99% when you count units delivered and 92 to 94% when you count order lines. Same orders, same months. Neither number is wrong. They answer different questions, and a retailer's scorecard may ask a third one.
The report shows both, labeled, so nobody has to guess which one the headline card means. Ask an AI assistant "what is our OTIF" against an unlabeled model and it picks a grain without telling you.
Kettle River BI was instrumental in helping us break through the hundreds of Excel-based data management solutions that managers had accumulated to provide answers to questions unique to their own departments, and replace them with flexible BI-based reports that facilitated team-oriented approaches to viewing data and solving problems.
Kettle River BI brought much more to the table than simply building Power BI reports. BI staff were the patient voice in the room, explaining how the tools worked, demonstrating how to manipulate them, and facilitating productive discussion between stakeholders while we were learning. Kettle River was truly transformational for our organization, both in the way we looked at data and the way we worked together.
Packaged food manufacturer, Pacific Northwest
Available as a reference for prospective clients. Ask us on the discovery call.
The Newsletter
Manufacturing Intelligence, in your inbox.
Real examples of the numbers manufacturers argue about: loss and waste, OTIF, inventory accuracy, AI readiness. How each one is calculated, where it goes wrong, and what to do about it. Free, and you can unsubscribe anytime.
You talk to Mitch.
Want similar results?
If your organization has similar challenges, let us explore whether a similar approach would work for you.