Your 8D lands in a customer quality engineer's inbox along with a dozen others. They won't read it like a novel, from D1 to D8. Most reviewers jump straight to the sections that tell them whether the supplier really understands the problem, and they form a first impression within minutes.
This post walks through that reading order from the reviewer's side, so you can check your own report the same way before you send it. It's a different angle from my earlier post on the 8 reasons customers reject supplier 8D reports: that one was a checklist by discipline; this one is about what gets looked at first and why. If you'd like an automated skeptical first read, there's a free pre-send readiness check, 8D Gate. More on it at the end.
Minute 1: Does D2 match the complaint?
The reviewer has their own complaint record open next to your report. The first check is simple: is this the same problem?
They compare part number, defect description, quantity, dates, and where it was found. If your D2 says "dimensional issue" and their complaint says "burr on bore edge, 37 pcs found at incoming inspection," they already suspect the rest of the report is generic. A sharp D2 with the customer's own numbers in it signals that you've read the complaint carefully.
Your check: Could the reviewer match your D2 to their complaint line by line, without guessing?
Minute 2: Are there real numbers in D3?
Next, containment. The reviewer wants to know whether any more bad parts are on their way, so they look for three things: quantities, locations, and a break point.
Here's an illustrative example (made up for this post, not from a real report):
Weak D3 (illustrative): "All stock was sorted and no further defects were found. Containment is effective."
Improved D3 (illustrative): "Sorted 100% by visual inspection under 10x magnification: supplier warehouse 2,400 pcs checked, 6 rejected; in transit 800 pcs checked, 0 rejected; customer site 1,150 pcs checked by our sort team, 3 rejected. First lot known good: lot 24-118, produced after tool change on [date]. All shipments from lot 24-118 onward are marked with a green dot."
The weak version asks the reviewer to take your word for it. The improved version gives them numbers they can check against their own stock, and tells them exactly which parts are safe.
Your check: Does D3 give checked and rejected counts at every location, the inspection method, and a clear break point the customer can use?
Minute 3: Does D4 explain both occurrence and escape, with evidence?
This is where most reviewers slow down. They look for two root causes, not one: why the defect was made (occurrence), and why your controls didn't catch it before it shipped (escape). And they look for evidence in the report, not just a statement.
Weak D4 (illustrative): "Root cause: operator did not follow the work instruction. Operator has been retrained."
Improved D4 (illustrative): "Occurrence: deburring tool insert worn beyond limit; tool life was not defined in the setup sheet, so inserts were changed only when an operator noticed. Confirmed by reproducing the burr with a worn insert and eliminating it with a new one (photos and measurements, attachment 3). Escape: final inspection checks the bore diameter with a plug gauge, which does not detect an edge burr; the control plan had no visual or tactile check for burrs. Confirmed by passing 5 known burred parts through final inspection, all accepted (attachment 4)."
"Operator error" tells the reviewer the analysis stopped early. The improved version says why the process allowed it, covers both causes, and shows how each was confirmed. More on both in How to write the D4 root cause in an 8D and Escape point in 8D.
Your check: Is there a separate occurrence and escape cause, each confirmed by data in the report?
Minute 4: Is there verification data in D5 and D6?
Reviewers scan D5 for whether each action actually addresses a D4 cause, then go straight to D6 looking for results. Not "actions implemented and effective," but data: how many parts were produced after the change, what was measured, what the result was, over what period.
If D6 is a single sentence, the reviewer usually stops there and writes back asking for evidence. If D6 still needs time to collect data, say what you'll measure and when, rather than claiming effectiveness early.
Your check: For every action, can the reviewer see what changed (D5) and the numbers showing it worked (D6)?
Minute 5: Did anything change in the system (D7), and does it all add up?
Finally, two quick checks that decide whether the reviewer reads the report as complete or not.
D7: Which documents were updated, with revision numbers? The control plan, PFMEA, setup sheet, work instruction? Were similar parts, tools, or lines checked? A D7 that names specific documents and other places the fix was applied tells the reviewer this won't come back next month on a sister part.
Consistency: Reviewers notice when the numbers don't agree. Quantities in D2 that differ from D3, dates where the action closes before the root cause was confirmed, a D4 cause with no matching D5 action, or an 8D marked closed while D6 says "monitoring ongoing." Each contradiction makes them doubt the rest.
Your check: Read the report top to bottom once looking only for numbers, dates and statuses. Do they all agree?
Reading your own 8D like a reviewer
The simplest habit: before you send, open the customer's complaint next to your 8D and follow the same five-minute path. D2 against the complaint, D3 numbers, D4 occurrence and escape with evidence, D5/D6 data, D7 and consistency. Anything you have to explain verbally is something the reviewer won't see.
If you'd like a second, skeptical reader, 8D Gate is a free pre-send readiness check. Upload an English 8D (PDF, DOCX or Excel) and in typically under a minute you get an Accept / Conditional / Reject advisory, a score, ratings per discipline, missing evidence, and the questions a reviewer is likely to ask. You review it on screen and can download a PDF.
Run the free 8D readiness check
8D Gate is a free pre-send readiness advisory. It doesn't replace the customer's decision. We don't store your 8D body in our database; you can delete anytime. Anthropic processes the text for the review.
Top comments (0)