Live data · rolling 30-day window

The State of CRM Copy

Across 29 anonymous scoring events in the last 30 days — real CRM operators, real drafts — here's what Orbit flagged, what scored, and what didn't.

Privacy: Orbit never stores the subject lines or drafts themselves. Only the score, the finding labels our own detector produces, and the character/word counts are kept — enough to build aggregate pictures, not enough to reconstruct anyone's copy.

Subject Line Scorer

11 subject lines scored

How they tier out

91%
9%
sharp (10)
decent (1)
risky (0)
spam (0)

Sharp = 85-100. Spam = sub-50. Anything below decent means the scorer caught at least one high-severity issue (typo, filler phrase, or shouting pattern).

Score distribution

85-100

10

70-84

1

Most-flagged issues

Missing space after punctuation

2

Subject and preheader could connect better

2

Borderline long for mobile — 56 chars

2

Contains "free" and 1 more

1

Preheader starts with lowercase

1

Subject trails off mid-thought

1

Borderline long for mobile — 54 chars

1

Labels come from Orbit's own detector — not user content. If "Content-free phrase in subject" is top, it means filler like "quick update" and "check this out" is still the single most common CRM-copy tell.

Slop Detector

18 drafts analyzed

How they tier out

78%
17%
sharp (14)
decent (1)
generic (0)
slop (3)

Sharp = 85-100. Slop = sub-50 — 3+ structural AI-drafting signals, typically uncompetitive with human-written copy.

Score distribution

85-100

14

70-84

1

0-49

3

Most-flagged slop patterns

Fake completeness

3

Corporate jargon

3

Generic closer

3

Anaphoric parallelism

3

Marketing adjective

3

Seamlessly adverb

3

Vague qualifier

3

Empty authority phrase

3

Fragment-heavy rhythm

2

Filler transition

2

Expect to see the template-marketing patterns dominate: "Let's dive into…", "In today's fast-paced digital landscape", empty authority phrases. These are the single clearest tell of AI-first-draft prose.

How this page updates

Every time someone scores a subject line in the subject-line scorer or runs a draft through the slop detector, Orbit logs an anonymous event: the score, the tier, and the detector labels that fired. The scored text itself is never stored.

This page reads those events across a rolling 30-day window and aggregates them. There are no curated case studies, no hand-picked examples, no weighting. What real operators score is what you see.

Want the monthly data drop by email? Start with the subject-line scorer and subscribe inline once you've got a result on screen.

Beyond the scorer

Get the full lifecycle toolkit inside Claude.

The subject-line scorer is one of 13 tools. Orbit for Claude packages all of them plus native Braze integration and 63 lifecycle skills into a single Claude Desktop extension. Free for everyone.