Step 1
A reply rate is read as a fraction; the stated denominator, not the percentage, decides whether a comparison is honest.
Public tool method · Version 1.0.0
Denominator Decoder
Use this method when you need to answer: What does this cold email reply-rate number actually mean, and is it supported by dated evidence? It turns the evidence you supply into a Sourced verdict and denominator-normalized comparison by applying explicit, versioned decision rules. The result is not a benchmark or certification. It shows the strongest evidence, riskiest input, next action, owner, and review date so another reader can inspect and challenge the decision.
Last material review 2026-08-08 · Next review 2026-11-06
Question answered
What does this cold email reply-rate number actually mean, and is it supported by dated evidence?
The Denominator Decoder re-expresses a reply-rate claim on a stated denominator and checks it against dated benchmark bands, using one versioned claims dataset that also generates the checker and the Claims Index.
Step 2
A total-sent or delivered denominator is placed on the same axis as the sourced benchmark bands; an opens or unique-opener denominator is refused, not converted, because Apple Mail Privacy Protection makes the conversion unverifiable.
Step 3
A number with no stated denominator returns the question to ask, treated as a first-class result rather than a failure.
Step 4
Every verdict and benchmark is a human-reviewed record with a primary source and date; material vendor disclosures trigger review and the active dataset receives a full review at least quarterly.
Step 5
Record the required inputs and label missing or estimated evidence before classification.
Assumptions and limits
- The decoder normalizes and compares stated numbers; it does not verify the sender's underlying data or predict campaign performance.
- Benchmark bands describe the populations their sources measured, which are not always B2B sales.
Usefulness hypothesis
A denominator-normalized, sourced comparison replaces denominator-free reply-rate claims in real outbound decisions.
Primary metric: Decoder result view, verdict copy, or Claims Index deep-link
Worked example and modes
The same method, preserved entry intent.
- Id
- reply-rate-single-number
- Claim Text
- The cold email reply rate is X%.
- Verdict
- MYTH
- Real Finding
- There is no single verified figure. Belkins measured 0.45% replies per total send across 7,530,489 emails in 2025. Instantly measured 3.43% per send the same year. Different denominators, different populations, both real.
- Do Instead
- Ask what the number is divided by before you repeat it. Report your own reply rate as replies over total sends, with the sample size and date range next to it.
- Category
- reply-rate-math
Distribution package
Use, embed, and share the instrument.
Blank CSV template90-second demonstrationThree share treatmentsPlain-language answer
Embed with canonical attribution
<iframe src="https://petrichorgrowth.com/tools/cold-email-benchmark-claims/embed.html" title="Denominator Decoder" loading="lazy" width="100%" height="900"></iframe>
<p>Source: <a href="https://petrichorgrowth.com/tools/cold-email-benchmark-claims">Denominator Decoder by Petrichor</a></p>Ownership and change control
Who keeps the result defensible?
Product
Petrichor product
Method
Petrichor strategy
Evidence
Petrichor research
Privacy
Petrichor operations
Security
Petrichor web platform
Distribution
Petrichor growth
Support
Petrichor client experience
A material method change requires a reason, impact note, fixture rerun, semantic-version increment, and public entry in the method change log.
Related decision methods
Continue through the evidence chain.
AI Skill Assurance and Deployment Readiness
Determine whether a bounded AI workflow has the evidence and controls required for qualified review.
Authority Ledger
Identify where authority is accumulating, drifting, or going uncited.
Category Evidence Builder
Determine whether a proposed category has observable demand and usable evidence.
Related plain-language answers
AI Skill Assurance answerAuthority Ledger answerCategory Evidence answer
Method questions
What another reader should know.
What does the Denominator Decoder method produce?
A versioned Sourced verdict and denominator-normalized comparison with visible reasoning, evidence status, next action, ownership, and review date.
What inputs does Denominator Decoder require?
Only the structured decision evidence named by the instrument. Missing or estimated inputs remain visible in the result.
Does the method use AI or a cross-customer benchmark?
No. It applies deterministic rules to the user's inputs and does not compare one customer with another.
When is the method reviewed or changed?
The current method is reviewed at least every 90 days. Material changes require new fixtures, a version increment, and a public change note.
When should the artifact move into facilitated work?
When the unresolved decision requires team challenge, evidence reconciliation, dissent, ownership, and an organizational commitment.