# Work-in-Process Visibility for $40M Manufacturers Canonical: https://granular.to/blog/wip-visibility-mid-market-manufacturers Published: 2026-05-27 Updated: 2026-05-27 Author: Trey Category: Playbook Tags: manufacturing, operations, automation, custom-software > A 90-day, $15-40K WIP visibility playbook for $30-75M manufacturers that fits between paper travelers and a $250K full MES rollout, focused on barcode scanning, cell-level tablets, and one dashboard that actually gets used. > **TL;DR.** 70% of mid-market manufacturers still collect shop floor data manually, and most lose 5-8% of revenue to invisibility: missed ship dates, idle WIP, working capital trapped in piles of parts nobody can find. The fix isn't a $250,000 MES. It's three barcode stations, one dashboard nobody currently has, and 90 days of operator discipline. Here's the playbook. Your foreman walks the floor at 7am and writes down where every job is. By 11am his notes are stale. At 2pm the office calls asking about Job 2847 and the answer is "I'll have to walk and find out." That's not a process problem. That's the absence of one. This is what work-in-process invisibility actually looks like at a $40M manufacturer. It costs you money in three ways that nobody on your floor will articulate: missed ship dates that lose repeat orders, working capital trapped in WIP nobody can find, and decisions made on Monday night data when it's already Tuesday afternoon. The good news: you don't need a Manufacturing Execution System to fix it. You need a middle answer between paper travelers and a $250,000 vendor rollout. Here's what that actually looks like. ## What WIP Invisibility Actually Costs You Most articles about WIP visibility talk about throughput, lean manufacturing, and data-driven decision-making. That's not where the money is. The real cost is working capital. According to [Descartes Finale's inventory analysis](https://www.finaleinventory.com/accounting-and-inventory-software/carrying-cost), inventory carrying costs run 20-30% of inventory value annually. Every week a job sits in WIP, you're paying roughly 0.4-0.6% of its cost just to hold it. A $40M manufacturer with $4M in work-in-process and a 14-day average cycle is bleeding $80,000 to $120,000 a year on WIP that exists only because nobody knows where it is or what to do with it. That's before you count the deeper costs: - **Missed ship dates that lose repeat orders.** Promise Tuesday, ship Friday, lose the rebid next quarter. - **Hot-job heroics.** Foreman drops what he's doing every time the office calls to find a job. Three of those a day adds up to a full day of supervisor time per week, just for search. - **Stale scheduling.** Operators backfill labor and quantity at end of shift. Tuesday morning's schedule decisions are based on Monday's reality. - **Phantom capacity.** Quoting assumes 80% utilization. Real utilization is 62% because two cells are choked with WIP nobody routed. [CAI Software's MES implementation analysis](https://caisoft.com/resources/mes-implementation-steps/) puts the total cost of running blind at 5-8% of annual revenue for manufacturers without proper shop floor systems. At $40M, that's $2M to $3.2M a year leaking through cracks you can't see because the data is on someone's clipboard. ![Stack of paper job travelers in a parts tote at a manufacturing workcenter showing the lost WIP visibility problem](/images/blog/wip-visibility-mid-market-manufacturers-travelers.jpg) ## What "Visibility" Means at a $40M Shop The MES vendor pitches start at $250,000 and assume you have an IT department. You don't. Your IT person is the office manager's nephew, and he changes the printer ink. What you actually need is the middle answer: enough instrumentation to know where every job is in real time, light enough that your foreman doesn't fight it, and cheap enough to fund out of a single quarter's freed-up working capital. That middle answer has three parts. ### 1. Barcode stations at every workcenter One tablet per cell. Mounted, not handheld. Operators scan a job traveler in when work starts and scan out when it's done. Two taps maximum. If a transaction takes more than 5 seconds, your operators will route around it within 30 days. This sounds primitive. It is. It also gives you something you don't currently have: a real-time record of where every job is, who's working on it, and how long it's been in the cell. Per the [Manufacturing Leadership Council](https://manufacturingleadershipcouncil.com/seventy-percent-of-manufacturers-still-enter-data-manually-2-37141/), 70% of manufacturers still enter shop floor data manually. That number isn't a tech adoption story. It's a usability one. The digital alternatives most vendors sell are worse than the clipboard. A tablet with a 12-second login that crashes once a shift is genuinely worse than a paper traveler. The bar to beat is "the clipboard worked." Your scan stations need to be faster than the clipboard or they'll join the dead tablets in the corner. ### 2. One dashboard nobody currently has Pull the scan data into one view: WIP by workcenter, aged greater than 48 hours. That's it. That's the dashboard. Not a pretty BI suite. Not a real-time digital twin. A single screen, TV on the wall of the office, or a tab on the scheduler's monitor, that shows every job in the building, where it is, and how long it's been sitting. This is the answer to the "I'll walk and find out" call. The office sees the answer before they pick up the phone. ### 3. A weekly aged-WIP review Every Monday morning, your scheduler and foreman look at the dashboard together for 15 minutes. Every job over 48 hours in a workcenter gets a single question: "Why is it still here?" The answers are almost always operational: waiting for a part, waiting for an outside operation, hot job got jumped, setup got pulled. The act of asking the question on a recurring cadence is what changes behavior. Aged WIP that nobody asks about stays. Aged WIP that gets asked about every Monday goes away within six weeks. ## Where Most MES Rollouts Die If the answer is this simple, why do most mid-market MES projects stall? > **A plant manager at a $55M precision machining shop, paraphrased from a recent conversation:** "We spent $180,000 on an MES. After six months, the foreman was still using the clipboard. The screens were slow, the login kept timing out, and the data we needed at the cell never showed up at the cell. It went to the office and the operators never saw it. So they stopped entering it." Three failure modes, in order of frequency: **Operator experience is too heavy.** Anything more than two taps per transaction dies. Vendors love rich operator interfaces because they justify a higher price. Operators love whatever lets them get back to making parts. **Data flows up but not down.** The dashboard sits in the office. The operator entering data never sees the result. So they stop seeing the point. **Integration with the ERP is broken.** Job numbers in the MES don't match job numbers in the ERP. Labor data has to be re-keyed. Scheduling has to be manually reconciled. The MES becomes a parallel system instead of a layer. Build for the operator first. The CFO's report can come later. ![Shop office monitor displaying a simple aged work-in-process dashboard by workcenter for a mid-market manufacturer](/images/blog/wip-visibility-mid-market-manufacturers-dashboard.jpg) ## The 90-Day, $15-40K Playbook Here's what a $40M manufacturer can actually execute this quarter: **Weeks 1-2:** Walk the floor and count travelers older than 5 days. If you can't find them in 30 minutes, you've quantified your hidden WIP. Document the workcenters with the most lost time. Usually three to five cells generate 80% of the visibility problem. **Weeks 3-6:** Install barcode stations at the worst three workcenters. Use whatever scanner pairs with your existing ERP. If you're on JobBOSS, Global Shop Solutions, or Epicor Kinetic, scan-in/scan-out modules already exist. Train operators in 20-minute sessions, one cell at a time. Two taps maximum. Mount the tablets so they don't walk off. **Weeks 7-10:** Build the dashboard. WIP by workcenter, aged greater than 48 hours, refreshed every 15 minutes. If your ERP has a half-decent reporting layer, this is a 4-hour build. If it doesn't, a custom dashboard wired to the ERP's transaction tables runs $8,000 to $15,000 from a competent shop. Either way, ugly and right beats pretty and late. **Weeks 11-12:** Start the Monday WIP review. Show up. Ask the question. Don't skip it. If you outgrow the dashboard at 18 months and graduate to a real MES like [Tulip](https://www.gartner.com/reviews/product/tulip-platform) or [MachineMetrics](https://www.g2.com/products/machinemetrics/reviews), you'll do it with three cells already instrumented, your operators already trained on scan discipline, and your scheduler already running off live data. The MES becomes a faster version of what you're already doing. Without the 90-day foundation, the MES is a $250,000 expression of hope. ## Why This Is a Working Capital Question, Not a Productivity Question The pitch your MES vendor is making (throughput, OEE, utilization) is the wrong pitch for a mid-market shop. The right pitch is working capital. Every week of WIP you free up converts directly to cash. A $40M manufacturer with a 14-day WIP cycle that drops to 10 days frees roughly $1.1M of working capital. At a typical 8% cost of capital, that's $90,000 a year. The barcode stations pay for themselves in the first month. Frame the project to your CFO as cash conversion, not productivity. The Deloitte [2025 Smart Manufacturing Survey](https://www.deloitte.com/us/en/insights/industry/manufacturing/2025-smart-manufacturing-survey.html) shows 41% of manufacturers prioritizing factory automation and 34% prioritizing active sensors over the next 24 months. Most of those investments are framed as productivity plays. The ones that actually get funded are the ones framed as capital plays. ## What This Looks Like Done Right A $50M aerospace machining shop ran a version of this playbook last year. Three cells instrumented. One dashboard. Weekly review. Six months in: average WIP cycle dropped from 16 days to 11. Hot-job calls to the foreman dropped by half. The dashboard caught two jobs that had been sitting at outside heat-treat for three weeks because the routing got missed. Total cost: $32,000 for hardware, integration, and a custom dashboard. Total working capital freed: $1.4M. Total time before they started talking to MES vendors about a "real" rollout: 18 months. The pattern is consistent. The middle answer works. ## FAQ **How long does it take to install barcode scan stations at three workcenters?** Hardware install runs 3-5 days per cell if the workcenter has power and Wi-Fi. Allow another 2 weeks for ERP integration testing and operator training. End-to-end, 4-6 weeks for three cells. **What if our ERP doesn't have a reporting layer that supports the dashboard?** Most mid-market ERPs (JobBOSS, Global Shop, Epicor Kinetic, IQMS) have a transactions or work-order activity table that a custom dashboard can wire into directly. Budget $8,000 to $15,000 for the build. If your ERP is older or genuinely closed, a small middleware layer that polls the ERP and writes to a separate dashboard database runs $15,000 to $25,000. **Will our operators actually scan?** They will if the scan takes under 5 seconds and they get something useful in return. The "something useful" is usually a status-by-cell view on the same tablet, so the operator can see the queue feeding their workcenter without walking to the scheduler. **Should we just buy a real MES instead?** Not yet. A $250K MES rollout at a $40M shop has a 50% chance of stalling within the first year, usually because the operator-experience problems we described above weren't fixed before the rollout. Spend $30K on the middle answer first. If you outgrow it in 18 months, the MES rollout will land cleanly because you'll already have the foundation. **What about AI on the shop floor?** AI helps once the data exists. Computer vision counting parts in a cell, voice-to-text status updates from operators, anomaly detection on scan patterns. These are real and valuable. They are also year-two problems. Year one is getting the scan data into a place where AI can read it. If you're running a $30-75M manufacturer and your foreman is still doing 7am walk-arounds, you don't need a $250K rollout. You need three scan stations, one dashboard, and a Monday meeting. That's the problem we [build for at Granular](/): fixed price, four weeks, working tool. If your Tuesday includes finding Job 2847 the hard way, [book 30 minutes with us](/) and we'll scope what the middle answer looks like at your shop. --- ## Keep Reading - **[Your ERP Says You Have 200 Units. You Have 147.](/blog/erp-says-200-units-you-have-147)**: Why ERP inventory counts go stale and how mid-market manufacturers reconcile what the system says with what's actually on the shelf. - **[How a $40M Manufacturer Captured 25 Years of Expertise](/blog/manufacturer-captured-25-years-expertise)**: A case study on capturing the tribal knowledge that lives in your senior operators' heads before they retire.