Every case on YourOutreachSUCKS is scored using a documented, weighted system. This page explains exactly how it works - the offense taxonomy, the scoring weights, and the standards that govern what counts as an offense.
Citing this methodology: If you reference the YourOutreachSUCKS scoring system in published work, see the citation formats at the bottom of this page. The taxonomy and scoring weights are original to this site and have been in use since February 2026.
Every case follows the same process from intake to publication. No step is skipped. Scores are computed mechanically from the offense weights - there is no editorial override of the final number.
1
Intake & redaction
Submission received. All identifying information removed immediately. Case number assigned. Redacted message stored; original destroyed.
2
Offense identification
Every phrase in the message is evaluated against the offense taxonomy. Each match is flagged as a candidate offense. A phrase can trigger multiple offense types simultaneously.
3
Severity classification
Each candidate offense is classified as Auto-fail, Critical, Major, or Minor. Classification follows the published definitions - not editorial judgment about the message's overall quality.
4
Score computation
Base score of 10.0. Each offense deducts its weight. If any Auto-fail offense is present, the score is capped at 2.0 regardless of total deductions. Final score rounded to one decimal place.
5
Grade assignment
Score converted to letter grade. A: 8.0-10.0 · B: 6.0-7.9 · C: 5.0-5.9 · D: 4.0-4.9 · F: below 4.0
6
Rewrite & publication
A rewritten version is drafted showing exactly what a better message looks like and why each change was made. Full audit published with evidence, scoring breakdown, analysis, and rewrite.
Scoring formula
Final Score= min(10.0- Σoffense_weights, auto_fail_ceiling) auto_fail_ceiling=2.0 if MERGE-FAIL or AI-WRITING present, else 10.0 Grade= f(Final Score)
Scores are floored at 0.1 - a message cannot score exactly 0 unless it contains no recoverable content. The lowest score on record is 0.1/10 (Case #0057 - THE ESSAY NOBODY ASKED FOR).
// Taxonomy
The offense taxonomy - origin and criteria
The taxonomy was built inductively, not deductively. No offense type was defined in advance. Each was added only after appearing independently in three or more cases, demonstrating it is a pattern rather than an isolated mistake.
The inclusion criterion is behavioral, not aesthetic: does this pattern meaningfully reduce the probability that the message achieves its goal? Style preferences are not offense types. Structural failures that undermine the message's core function are.
OffenseCategoryCasesAdded
FAKE-RESEARCHOpening31Feb 2026
PREMATURE-ASKClose30Feb 2026
VAGUE-ICPTargeting26Feb 2026
BUZZWORD-LOADExecution31Feb 2026
MERGE-FAILExecution5Feb 2026
HUMBLE-BRAG-PIVOTOpening4Feb 2026
FOMO-BAITClose34Feb 2026
PAIN-ASSUMPTIONOpening27Mar 2026
REPLY-ALL-ENERGYClose21Feb 2026
AI-WRITINGExecution5Mar 2026
WALL-OF-TEXTExecution1Feb 2026
// Weights
Scoring weights by severity
Each severity tier has a defined point range. Within the range, the specific deduction for an instance reflects the severity of that particular manifestation - a barely-visible merge tag deducts less than a [FIRST NAME] in the opener of a professional email.
SeverityRangeDefinitionExample
Auto-failCap at 2.0Makes the message undeniably broken or dishonest regardless of any other content quality. No recovery is possible.MERGE-FAIL with visible bracket
Critical-3.0 to -4.0Structural failure that undermines the core premise of the message. The message cannot work as intended with this offense present.FAKE-RESEARCH, VAGUE-ICP
Major-1.5 to -2.5Tactical error that meaningfully reduces effectiveness. The message can partially function but is significantly impaired.PAIN-ASSUMPTION, FOMO-BAIT
Minor-0.5 to -1.0Style or polish issue that compounds other problems. Rarely decisive alone but consistently weakens already-impaired messages.Excessive length, weak subject
Multiple offense instances of the same type stack up to a cap of 2x the single-instance weight. Repetition makes things worse, but there is diminishing return on tagging the same failure twelve times in a single message.
// Standards
Editorial standards that govern every case
S1
Offense, not taste
An offense must have a defensible causal mechanism connecting it to reduced effectiveness. "I don't like this phrasing" is not an offense. "This phrase signals to the reader that no research was done, because it contains no verifiable specifics" is.
S2
Three-case threshold for taxonomy inclusion
No offense type is added to the taxonomy until it has appeared independently in at least three separate cases. This prevents overfitting to a single unusual example and ensures the taxonomy reflects genuine patterns.
S3
Structural critique, never personal
Every critique is directed at the message structure, not the sender. "This phrase fails because it assumes interest that hasn't been built" - not "this sender is lazy." The sender is redacted. The pattern is the subject.
S4
Score immutability after publication
Once a case is published, its score is not retroactively changed if the taxonomy weights are updated. Scoring changes apply to future cases only. The changelog documents every scoring methodology update.
S5
Good outreach is documented too
Messages scoring B or above are published in the Good Outreach section. The taxonomy is not constructed to always find failures - messages that exhibit genuine research, specific ICP targeting, and customer outcomes score well and are recognized as such.
// Coinage
The taxonomy terms are original to this site
The offense taxonomy, all 30 named offense types including FAKE-RESEARCH, VAGUE-ICP, MERGE-FAIL, AI-WRITING, BUZZWORD-LOAD, PREMATURE-ASK, HUMBLE-BRAG-PIVOT, FOMO-BAIT, PAIN-ASSUMPTION, WALL-OF-TEXT, REPLY-ALL-ENERGY, and the v2.0 additions like CLICKBAIT-SUBJECT, DESPERATION-SIGNAL, FALSE-FAMILIARITY, CTA-OVERLOAD, CHANNEL-DEAF, and more, is original to YourOutreachSUCKS. These terms were coined here to name patterns that existed without names.
On coinage and citation
These terms are coined by YourOutreachSUCKS (youroutreachsucks.com) and have been in documented use since February 2026. If you use these terms in published work - articles, research, training materials - we ask that you attribute them to this site. They are not trademarked. They are simply original, and attribution helps maintain the integrity of the definitions.
The definitions on this site are the canonical versions. If these terms appear elsewhere with different definitions, defer to the definitions published here unless a deviation is explicitly noted with a reason.
// Citation
How to cite this methodology
If you reference the YourOutreachSUCKS scoring system, taxonomy, or individual offense definitions in published work, use one of the formats below.
Citation formats
APA
Gierlich, J. (2026). YourOutreachSUCKS offense taxonomy and scoring methodology (v2.1). YourOutreachSUCKS. https://youroutreachsucks.com/methodology
It's a 6-dimension rubric for grading cold outreach. Messages are scored on personalization, relevance, value proposition, CTA quality, tone, and execution. Each gets a 0-10 score, and the weighted average produces the overall grade.
How are cold emails scored on YourOutreachSUCKS?
Six dimensions, each weighted. The weighted average is the final number. Right now the average across all filed cases is 5.4/10.
What are auto-fail rules in cold email scoring?
Two offenses cap the score at 2.0/10 no matter what else the message does right. MERGE-FAIL (visible template variables like [First Name]) and AI-WRITING (obviously machine-generated copy).
What is the cold outreach offense taxonomy?
30 named failure patterns. Each one describes something specific that keeps showing up in bad outreach. Every term was coined by YourOutreachSUCKS after the pattern appeared across multiple documented cases.