The Glenday sieve is a method for sorting products by volume to find the small share of SKUs that drives most of your output. You rank every product by volume, add up cumulative volume, and cut the list into bands at the 50, 95, 99, and 100 percent points. The top band, your runners, is where leveled, fixed-cycle production starts.
The method comes from Ian Glenday, who noticed the same pattern across plant after plant when he sorted products by volume. Because the 80/20 idea was already named for Vilfredo Pareto, Glenday named his version the Glenday sieve (Lean Enterprise Institute, Levelled Production: Where to Start). The sieve answers a specific question: given a messy, high-mix product list, where do you start leveling? The honest answer is almost never "everywhere at once." It is "start with the handful of products that make up half your volume," and the sieve is how you find them. It is a practical planning tool inside lean manufacturing and it pairs directly with production leveling.
What Is the Glenday Sieve?
The Glenday sieve is a sorting technique, not a chart type. You take your product list, rank it from highest volume to lowest, and calculate the running cumulative share of total volume as you go down the list. Then you draw lines where the cumulative total crosses 50 percent, 95 percent, 99 percent, and 100 percent. Those four cuts create four bands, usually color-coded green, yellow, blue, and red. What makes the tool striking is how lopsided the bands turn out to be. Glenday's analyses consistently landed near the same ratios: about 6 percent of products give 50 percent of the volume, roughly 50 percent of products give 95 percent, and around 70 percent give 99 percent. The remaining 30 percent of the product list, often a huge count of SKUs, accounts for the last one percent of what you actually build.
What Are the Four Bands?
Each band gets a color and a role, and the roles are what tell you how to plan each group. The classic lean names are runners, repeaters, and strangers, with the last band being the long tail.
| Band | Cumulative volume | Typical share of SKUs | Role and how to run it |
|---|---|---|---|
| Green | First 50% | ~6% | Runners. High, steady volume. Put these on fixed, repeating cycles first. |
| Yellow | 50% to 95% | ~50% cumulative | Repeaters. Regular but smaller. Schedule on a looser recurring pattern. |
| Blue | 95% to 99% | ~70% cumulative | Strangers. Infrequent. Fit into planned capacity slices, not the fixed cycle. |
| Red | Last 1% | The remaining tail | The long tail. Make to order, review for rationalization, or batch rarely. |
The percentages are Glenday's typical findings, not a law, your plant will land somewhere near them, not exactly on them. The value is not the precise numbers but the shape of the answer: your green runners are few, and they are where leveling pays off fastest.
How Do You Build a Glenday Sieve?
The mechanics are simple enough to do in a spreadsheet in an afternoon. The discipline is in acting on what it shows.
- Pull a clean volume list. Every SKU with its volume over a representative period, often 6 to 12 months so seasonality does not distort the ranking. Use volume, or value if that is the better measure for your business.
- Rank high to low. Sort the list by volume, largest first. This ordering is the whole trick; everything else follows from it.
- Add a cumulative percentage column. Running down the list, total the volume and express it as a percent of the grand total at each row.
- Cut at 50, 95, 99, and 100 percent. Draw the band lines where cumulative volume crosses each threshold. The row counts in each band are usually the surprise.
- Color-code the bands. Green, yellow, blue, red. The colors make the sieve legible to planners and operators at a glance, which is the point.
- Name the runners and level them first. Take the green band and design a fixed, repeating production cycle for it before touching anything else.
- Decide a policy for each other band. Repeaters on a looser cycle, strangers into planned capacity, the red tail reviewed for make-to-order or rationalization.
How Does the Sieve Enable Fixed-Cycle Production?
This is why Glenday built the tool: leveling a high-mix plant looks impossible until the sieve shows you that most of the mix does not need leveling. The green runners are high-volume and steady, which means they can run on a fixed, repeating cycle, the same products in the same sequence, week after week. That fixed cycle is exactly what a heijunka box executes, and it is the practical on-ramp to production leveling. You do not try to level 400 SKUs. You put the 20 or 30 runners on a repeating wheel, handle repeaters on a looser rhythm, and treat strangers and the tail as planned exceptions. The plant gets the stability benefits of leveling where they matter most, without pretending every low-volume SKU can ride the same cycle. A stable runner cycle also makes standard work hold, because the pacemaker stops lurching between wildly different products.
How Is the Glenday Sieve Different From a Pareto or ABC Analysis?
They are cousins, and the difference is in purpose. A Pareto chart ranks causes or categories to find the vital few, usually to prioritize problem-solving. A classic ABC inventory analysis segments SKUs by value to set stocking and counting policy. The Glenday sieve uses the same rank-and-cumulate move but aims it at a production question: which products are stable and high-volume enough to level on a fixed cycle? The four cuts at 50/95/99/100 and the runner-repeater-stranger language are tuned for scheduling decisions, not inventory control or defect analysis. In practice you might run all three, but the sieve is the one that tells you where leveled flow can start.
Two mistakes sink a sieve. The first is treating Glenday's ratios as targets to hit rather than patterns to observe, your plant lands where it lands, and forcing the data to match 6 percent is just fooling yourself. The second is building the sieve once and never refreshing it. Product mix drifts as customers, seasons, and launches change, so a runner today can slide into the repeaters next quarter. Rerun the sieve on a regular cadence, and treat a product's movement between bands as a signal to change how you schedule it.
By the Numbers: How Concentrated Volume Really Is
The sieve keeps working because real product lists are far more concentrated than most teams assume. Glenday's repeated finding, published through the Lean Enterprise Institute, was ratios near 50/6, 95/50, and 99/70, roughly 6 percent of products producing half the volume, half the products producing 95 percent, and 70 percent producing 99 percent (Lean Enterprise Institute, Levelled Production: Where to Start). The practical consequence is large: you can capture most of the stability and inventory benefit of leveling by fixing the cycle on a small green band, rather than boiling the ocean across every SKU. And the payoff shows in inventory, U.S. Census Bureau data puts manufacturers' inventories-to-shipments ratios around 1.4 to 1.5 in recent years (U.S. Census Bureau, M3 Survey), much of it batch-driven stock that a leveled runner cycle directly reduces.
Where Harmony fits: a sieve is only as good as the volume data behind it, and that data usually lives scattered across an ERP, spreadsheets, and a scheduler's memory. Harmony connects machines, systems, and paperwork into one real-time operational layer so the SKU-by-SKU volumes you sieve are pulled from what the floor actually ran, and the bands stay current as demand drifts instead of freezing at last year's snapshot. See what that looks like in a plant like yours in the CLS case study no rip-and-replace to get there.