Runner, repeater, stranger (RRS) is a classification method that sorts every product a plant makes into three buckets based on how often and how predictably it is ordered, so that scheduling, cell design, and inventory policy can be matched to each demand pattern.
The idea is simple: not every SKU deserves the same treatment. High-frequency, stable "runners" reward dedicated flow, while erratic "strangers" need flexible capacity and different planning rules. Get the classification right and you stop forcing a single scheduling logic onto a mixed portfolio that will never behave uniformly.
The labels describe a spectrum of demand frequency and predictability, not size or complexity. A part can be physically small and still be a stranger if nobody can forecast when it will be ordered.
RRS pairs naturally with an ABC analysis of demand. Where ABC ranks items by value or volume contribution, RRS layers on the frequency and variability dimension that tells you how to schedule them.
Suppose a machining plant runs 40 part numbers over a 250-day working year. You pull two data points per part: how many days it was ordered (order frequency) and how much of total annual volume it represents.
The pattern is the classic long tail: 6 runners drive two-thirds of throughput, while 20 strangers eat setup time and planning attention for a sliver of output. A Pareto analysis of this same data makes the imbalance visible at a glance and tells you exactly where dedicated flow will pay off.
Once parts are classified, the shop floor layout follows. Runners justify dedicated or near-dedicated cells with balanced, standardized flow because their steady demand keeps that capacity busy. Repeaters typically live in shared, quick-changeover cells that a family of parts rotates through. Strangers are best served by a flexible "job shop" area or a general-purpose cell that absorbs variety without disrupting the runner lines.
This separation protects your best flow from your worst variability. When strangers are forced through a runner cell, every low-volume changeover steals capacity and injects instability. Before committing to a layout, map the current routings with a value stream map and check operator and material travel with a spaghetti diagram so you cut motion waste rather than freeze it into new cells.
EPEI (Every Part Every Interval) is the cadence at which a shared resource cycles through all the parts assigned to it. RRS is what lets you set a realistic EPEI, because you only need a tight rhythm for the parts that actually repeat.
Consider a press cell with 6.5 hours of available run time per shift and 90 minutes of that consumed by changeovers. If you assign 5 repeaters and each changeover takes 18 minutes, five changeovers fit inside that 90-minute budget, giving an EPEI of one day: every part gets made every single day.
Reduce changeover time through constraint-focused improvement and you can either add parts to the interval or shorten it. Strangers stay out of this rhythm entirely; they are slotted into open capacity or a dedicated flexible window so they never destabilize the repeater cadence.
A shorter EPEI means smaller batches, lower inventory, and faster response, which is the core promise of a pull system replenished with kanban. But it only works if changeover capacity supports it. That is why RRS and EPEI are inseparable.
The classification also tells you where continuous flow beats batch production. Runners with stable demand are ideal candidates for level scheduling and single-piece or small-lot flow, because you can smooth them across the week without building large safety buffers. Repeaters usually run in modest batches sized by the EPEI.
Strangers, with unpredictable timing, are handled make-to-order and often need a small safety stock of raw material or components rather than finished goods you may never sell.
Inventory strategy shifts accordingly. Runners can carry finished-goods buffers sized by Little's Law and monitored through inventory turnover. Strangers should almost never sit as finished stock. Reclassify periodically, because a stranger that starts repeating deserves promotion, and a runner that fades should be demoted before it clogs a dedicated line.
RRS is only as good as the demand and performance data behind it, and that is where Fabrico serves as the real-time data foundation. Fabrico captures real-time OEE and production monitoring on every cell, so you can see actual run frequency, changeover time, and downtime per part rather than relying on stale spreadsheets.
Its computer vision works even on machines with no PLC, which means older equipment running low-volume strangers is measured too. Watching genuine OEE by product family tells you whether your runner cells are truly stable and whether changeover reduction is unlocking a tighter EPEI.
On the maintenance side, Fabrico is a field-ready CMMS with work orders, assets, preventive scheduling, and spare-parts tracking, so the cells that flow your runners stay available when the schedule depends on them. Fabrico is EU-built with EU data residency. Explore the OEE monitoring and CMMS capabilities to ground your RRS decisions in measured reality.
Review the classification on a regular cadence, typically quarterly, and always after a significant demand shift or new product introduction. Demand patterns drift: a stranger can become a repeater as a customer scales up, and a runner can fade toward the tail. Re-running the analysis on fresh, real production data keeps your cell assignments and EPEI honest rather than optimizing for last year's mix.
At minimum you need order frequency (how many periods each part was demanded) and volume contribution per part over a representative window, ideally a full seasonal cycle. Adding changeover time and demand variability sharpens the picture, because two parts with identical volume can behave very differently if one is steady and the other spiky. Real-time production monitoring makes these numbers trustworthy instead of estimated.
Yes. Runner status is about frequency and predictability, not physical size or annual quantity. A tiny component ordered every day in stable amounts is a runner and belongs in steady flow, while a large custom assembly ordered twice a year is a stranger. Classify by demand behavior, then let volume and value guide how much dedicated capacity each part earns.
Ready to classify your portfolio on real numbers instead of guesswork? Book a Fabrico demo and see how live OEE and production data reveal which products are runners, repeaters, and strangers, so you can design cells and set an EPEI that actually holds.
Login
Blog