How to Cut Unplanned Downtime by 40% in a Small Factory

The math that makes this worth your time

Before we get into the method, let's talk numbers. Say your factory has 40 hours of unplanned downtime per month across all equipment. Your fully burdened cost per downtime hour — lost production, idle labor, scrap, expedited shipping to make up late orders — is $300/hour. That's $12,000/month walking out the door.

A 40% reduction brings you to 24 hours of unplanned downtime. That's $4,800/month saved — $57,600 per year. The PM tasks and tracking system that achieve this cost a fraction of that. The ROI isn't theoretical; it's arithmetic.

The question isn't whether reducing downtime pays off. It's whether you can actually do it. The answer is yes — if you follow the data instead of guessing.

Step 1: Track downtime by reason code for 30 days

This is the hardest step because it requires discipline without immediate reward. For 30 days, every time a machine goes down unplanned, someone records three things:

1. Which machine — asset name or number
2. Why it stopped — the reason code (hydraulic leak, electrical fault, bearing failure, operator error, material jam, tooling break, etc.)
3. How long it was down — start time to restart time, in minutes

That's it. Don't try to fix anything differently during this period. Don't launch new PM programs. Just record what happens. You need clean baseline data, and changing your behavior during the measurement period contaminates it.

Keep the reason code list short — 8 to 12 codes maximum. If you give people 50 options, they'll pick “Other” every time. Good starter codes:

• Hydraulic/pneumatic failure
• Electrical/controls fault
• Bearing/motor failure
• Belt/chain/coupling failure
• Material jam or feed issue
• Tooling failure
• Lubrication-related
• Operator error
• Quality hold (process out of spec)
• Unknown / investigation needed

Make it dead simple to log. A clipboard on the machine, a QR code that opens a mobile form, or a whiteboard in the break room. The format doesn't matter as long as people actually fill it in.

Step 2: Pareto the data

After 30 days, sort your downtime data by total hours lost per reason code. In almost every factory, the Pareto principle holds: the top 3 causes account for 60–80% of total downtime hours.

This is the moment most small factories skip — and it's the most important one. Without this analysis, maintenance improvements are shotgun blasts. With it, they're rifle shots.

A real example: a 35-person metal fabrication shop tracked 45 hours of monthly unplanned downtime. The Pareto showed:

1. Hydraulic failures — 18 hours (40%)
2. Material jams — 9 hours (20%)
3. Electrical faults — 7 hours (16%)
4. Everything else — 11 hours (24%)

The owner assumed electrical problems were their biggest issue because those failures were dramatic — alarms, error codes, frantic calls to the controls tech. But hydraulic failures, which were quieter (a slow leak here, a pressure drop there), added up to nearly three times the downtime.

Step 3: Build targeted PMs for the top cause

Don't try to fix everything at once. Pick the number-one cause and build PMs specifically designed to prevent that failure mode.

In the example above, the shop dug into their hydraulic failures and found three recurring issues:

• Hose failures from heat degradation (hoses routed too close to heat sources)
• Pump wear from contaminated oil (filters not being changed on schedule)
• Cylinder seal leaks (seals drying out on machines that sat idle for weeks between jobs)

The targeted PMs they built:

• Monthly hydraulic hose inspection — visual check for cracking, bulging, and chafing. Replace any hose showing wear before it bursts. Re-route hoses away from heat sources.
• Every 500 hours: hydraulic filter change — no exceptions, no “it looks okay.” Just change it.
• Every 2,000 hours: hydraulic oil sample — send to a lab. Test for particle count, water content, and viscosity. Change oil when the lab says to, not on a fixed interval.
• Weekly: cycle idle machines — run hydraulic cylinders through their full range of motion once a week to keep seals lubricated, even on machines not in production.

Notice how specific these are. Not “do hydraulic maintenance.” Each PM targets a specific failure mode that showed up in the data.

Step 4: Measure the change over 90 days

Keep tracking downtime by reason code for the next 90 days — using the exact same method as your baseline period. Don't change the codes, don't change the process, just keep recording.

After 90 days, compare. The metal fabrication shop from the example above saw these results:

• Total monthly unplanned downtime dropped from 45 hours to 22 hours — a 51% reduction
• Hydraulic failures specifically dropped from 18 hours to 4 hours
• Material jams were unchanged (they hadn't addressed those yet)
• At $300/hour, that's $6,900/month saved — over $82,000 per year

The 90-day measurement period matters for two reasons. First, it confirms the improvement is real and sustained, not a fluke. Second, it gives you credibility with your team and your boss. “We reduced downtime by 51%” backed by data is far more powerful than “I think things are better.”

Then repeat for the next cause

Once the top cause is under control, move to number two on the Pareto. In the example, that was material jams. Investigate the specific failure modes, build targeted PMs, implement them, and measure again.

Each cycle gets you another chunk of downtime reduction. The improvements compound because you're systematically eliminating the biggest contributors first. After 2–3 cycles (6–9 months), you've typically addressed the causes that represent 70–80% of your original downtime.

Common mistakes that derail the process

Having watched dozens of small factories attempt this, here are the pitfalls:

• Trying to fix everything at once — the Pareto exists for a reason. Focus beats breadth.
• Skipping the baseline — “we know our problems” is usually wrong. The data regularly surprises even experienced maintenance managers.
• Building PMs that are too general — “inspect hydraulic system” is not a PM. “Check hydraulic filter differential pressure gauge and replace filter if above 15 PSI differential” is a PM.
• Not measuring the result — without the 90-day follow-up, you don't know if your PMs worked. And you can't justify the next round of investment.
• Stopping after one cycle — the first win is motivating. Use that momentum to tackle the second cause, then the third. This is a continuous improvement process, not a one-time project.

What your downtime really costs

Most factories underestimate downtime cost because they only count the obvious: lost production and repair parts. The full cost includes:

• Idle labor — operators standing around waiting for the machine to come back
• Scrap from the failure — the parts in process when the machine went down are often scrap
• Startup scrap — the first 15–30 minutes of production after a restart often produce out-of-spec parts
• Expedited shipping — you missed the delivery window, so now you're paying for overnight freight
• Overtime — running the weekend to make up lost production
• Customer impact — late deliveries erode trust. Enough of them and you lose the account.

When you add these up, $200–$500/hour of unplanned downtime is typical for a small factory. That makes the ROI on even modest downtime reduction substantial.

Start this week

You don't need to buy anything or implement anything to start Step 1. Print a simple log sheet, hang it on each machine, and tell your operators: “When a machine goes down, write the machine name, why, and how long.” That's it.

In 30 days, you'll have data. The data will tell you where to focus. The focus will produce results. And the results will pay for whatever system you use to manage the process going forward.