91Ó°ÊÓ

A product fails in the field, or on the line, and the instinct is to move fast: identify what broke, patch it, and get the line or the shipment moving again. That instinct is understandable and almost always the wrong first move. Speed without structure tends to produce a fix aimed at the symptom instead of the cause, and the failure comes back weeks or months later, sometimes worse than the first time.

The financial stakes make the discipline worth it. U.S. manufacturers paid , a 4% increase over 2024, according to Warranty Week’s annual product warranty report. Every dollar in that figure represents a failure that reached a customer, and most of those failures trace back to a root cause that a rushed investigation missed the first time.

This article covers the steps that belong at the beginning of a product failure investigation, before root cause analysis tools like fishbone diagrams and 8D reports even enter the picture.

---

Key Takeaways

• U.S. manufacturers paid $30.37 billion in warranty claims in 2025, up 4% from the prior year, a figure that reflects failures that already reached customers.
• The American Society for Quality reports that total quality-related costs reach 15% to 20% of sales revenue at many manufacturers, and most of that cost stays hidden in rework, re-inspection, and engineering time rather than showing up as visible scrap.
• The first two steps of any failure investigation are containment and evidence preservation, both of which happen before anyone starts diagnosing the cause.
• A precise, measurable problem statement is what separates a productive investigation from one that chases the wrong lead for weeks.
• Fishbone diagrams organize the field of possible causes; the 5 Whys technique drills into the most likely branch to find the cause underneath the symptom.
• The 8D methodology sequences a failure investigation from team formation and containment through root cause, corrective action, and prevention, closing the loop instead of stopping at the fix.

---

Why “Just Fix It” Backfires

Every manufacturer carries a cost of poor quality, and it’s larger than most teams assume. The American Society for Quality has found that at many manufacturers.

Scrap is the visible piece of that number. The larger, hidden piece sits in rework hours, re-inspection, engineering time spent chasing repeat problems, and the warranty and brand costs that show up after a defect ships. Industry estimates put hidden failure cost at roughly ten times the visible scrap cost once all of that is counted.

A rushed fix feeds that hidden cost directly. Quality management writing on the subject, including analysis published by senior quality professionals like Roxann Dawson of Costain, rather than a one-time manufacturing slip.

A design fault doesn’t go away because a single failed unit gets replaced or repaired. It resurfaces on the next unit, and the next, until someone traces the failure back to the actual design, material, or process decision that caused it.

---

Step One: Contain the Problem Before You Diagnose It

Containment comes first, ahead of any diagnosis. The goal is to stop the failure from reaching more customers or more units while the investigation is still underway. That can mean quarantining affected inventory, halting shipment of a specific lot, or pulling a supplier component from the line until its role in the failure is ruled out.

Containment decisions get made with incomplete information, and that’s expected. A team doesn’t need to know the root cause to decide that a batch showing a specific defect shouldn’t ship. What containment buys the investigation is time: time to collect evidence properly, time to interview the people who witnessed the failure, and time to test hypotheses without the pressure of a customer complaint arriving every day the investigation continues.

---

Step Two: Preserve the Evidence Before It Disappears

The failed part or unit is the most valuable piece of evidence in the investigation, and it’s also the easiest to lose. A well-meaning technician cleans it up, a production team scraps it to keep the line moving, or someone attempts a repair before anyone documents the original failure state. Any of those actions can erase the exact evidence a root cause analysis depends on.

The practical steps at this stage are straightforward. Photograph the failure from multiple angles before touching the part. Record the conditions under which it failed, including the process parameters, the operator, the shift, and the lot or batch number. Set the physical part aside in the condition it failed in, and establish a clear chain of custody for who has handled it since. None of this requires specialized tools. It requires treating the first hour after a failure is discovered as part of the investigation, not a cleanup task to get out of the way.

---

Step Three: Write a Precise Problem Statement

A vague problem statement sends an investigation in the wrong direction before it starts. “The part failed” or “customers are complaining” gives a team almost nothing to work with. A precise problem statement states what happened, in measurable terms, and separates the observed symptom from any assumption about the cause.

Root cause analysis guidance from Advanced Technology 91Ó°ÊÓ illustrates the difference well: a statement like gives an investigation team a specific, testable starting point. A statement like “the fill process is inconsistent” does not. The first version can be measured, replicated, and used to rule causes in or out. The second invites speculation.

---

Step Four: Assemble the Right Team

A single engineer working alone can chase a plausible-looking cause for days and still miss the real one, particularly when the failure sits at the intersection of design, materials, and process, which most product failures do. Bringing in a cross-functional team from the outset, rather than after an initial investigation stalls, is what keeps that from happening.

A functional failure investigation team typically includes the engineer or technician closest to the design or process in question, a quality professional who can interpret inspection and process data, and, when a purchased component is involved, someone with visibility into the supplier’s process. On more significant investigations, a facilitator trained in root cause methodology helps keep the team from settling on the first plausible explanation instead of the correct one.

---

Step Five: Map the Possible Causes, Then Drill Down

Once the problem is contained, the evidence is preserved, and the team is assembled, the investigation shifts to identifying the cause. Two tools, used together, cover most of this work.

A fishbone (Ishikawa) diagram organizes every plausible cause into categories, typically covering people, process, materials, equipment, and environment. Its value isn’t in identifying the single correct answer. It’s in making sure the team considers the full range of possibilities before narrowing in on one, which prevents the common mistake of fixating on the most obvious explanation while a less visible one goes unexamined.

The 5 Whys technique takes over from there. For each promising branch on the fishbone diagram, the team asks why the condition occurred, and then asks why again on that answer, typically five times, until the chain of reasoning reaches something the team can act on: a specific process step, a design tolerance, a training gap, or a supplier’s material specification. is one of the most common reasons a “corrected” failure comes back: the team stopped at a symptom that looked like a cause.

---

Where 8D and FMEA Fit In

Once a team has been through containment, evidence preservation, problem definition, and cause analysis, the 8D methodology provides the structure to close the loop formally. The eight disciplines run from team formation and problem definition through containment, root cause analysis, corrective action, implementation, and a final prevention step that asks where else the same failure mode could occur. That last step is where many investigations stop too early: fixing the specific instance of a failure without checking whether the same design or process condition exists on other products or lines.

FMEA (Failure Mode and Effects Analysis) works on the opposite end of the timeline. Where 8D investigates a failure that already happened, FMEA is a proactive exercise, run during design or process planning, that asks what could fail before a single unit is built. A mature FMEA process doesn’t replace the need for failure investigations, since no analysis catches every failure mode in advance, but a well-maintained FMEA does mean fewer investigations start from zero. Teams with a documented FMEA on file already have a list of anticipated failure modes to check the actual failure against, which shortens the path to root cause considerably.

---

From Root Cause to Corrective Action

Identifying the root cause isn’t the end of the investigation. The corrective action needs verification before a team declares the issue resolved, typically through testing or a monitored production run that confirms the failure no longer occurs under the same conditions that originally produced it. Skipping verification and moving straight to full implementation risks discovering, weeks later, that the fix addressed a contributing factor rather than the actual root cause.

The preventive step matters just as much. A root cause traced to a specific supplier’s material lot, a specific tolerance on a drawing, or a specific step in a work instruction is rarely isolated to the one product where it was discovered. Checking other products, lines, and suppliers for the same underlying condition is what turns a single failure investigation into a permanent reduction in cost of poor quality, rather than a fix that resolves one complaint and leaves the underlying condition in place everywhere else.

---

Where 91Ó°ÊÓ Can Help

A structured failure investigation draws on design engineering, manufacturing process knowledge, and supplier quality experience at the same time, and most internal teams don’t have all three represented in one room when a failure occurs.

At 91Ó°ÊÓ, we support manufacturers through manufacturing engineering and product development work that includes root cause investigation, corrective action planning, and the design and process changes that come out of it. Whether you’re in the middle of an active failure investigation or want to build a more structured process before the next one happens, we can help you get to the actual cause, not just the nearest fix.

Start the conversation here.