Media Outreach Benchmark Methodology
How AstraMedia plans, measures, and will publish findings about media outreach — and what we will and will not claim.
Why this page exists
This page explains how AstraMedia plans, measures, and will publish findings about media outreach — and what we will and will not claim. It exists so that every statistic we publish can be checked: what was counted, over what period, from what data, and by whom.
What this benchmark covers
AstraMedia's product tracks the outreach lifecycle: campaign created, pitch drafted, pitch edited, pitch approved, pitch sent, follow-up sent, response received, interview requested, coverage recorded, campaign closed. The Q4 2026 Media Outreach Benchmark is a set of descriptive statistics drawn from anonymized, aggregated product activity across real campaigns.
The first question set we intend to answer (each with the caveats below):
- Which media categories respond most often?
- How much does human personalization/editing of a pitch affect response rates?
- Which story angles perform best for founders vs consultants?
- What is the median time from story to first response?
- Which channels produce the most meaningful conversations?
- What causes non-response (targeting, timing, angle, message)?
Definitions (what counts as what)
Every outcome word is defined before counting — the same definitions are used in case studies and in the product event model:
- Pitch sent: a pitch message actually sent through a supported channel (email, social, podcast platforms, HARO-style).
- Response: a substantive, human reply from a journalist, outlet, host, or target contact. Excludes bounces, auto-replies, out-of-office, unsubscribe notices, and spam/filter noise. A polite decline counts as a response.
- Conversation: a multi-turn exchange — at least two substantive reply rounds between the same customer and the same contact about the story.
- Interview requested: the contact explicitly requests, agrees to, or schedules an interview, podcast appearance, comment, or other direct engagement.
- Coverage: published, verifiable output referencing the customer's company, name, or launch (article, mention, podcast episode, newsletter feature, broadcast segment).
Data standards (the metadata block)
Before publishing any statistic, AstraMedia records:
- Sample size — N for every denominator (campaigns, pitches sent, contacts reached)
- Date range — exact collection window in UTC
- Inclusion/exclusion rules — what counted and what was filtered (e.g. bounces and test campaigns excluded)
- Customer-level vs campaign-level — the unit of analysis, never mixed
- Known biases — including selection bias, small-sample volatility, and channel skew
- Confidence intervals — 95% CI or margin of error where N supports it; ranges instead of false-precision points for small N
- Named author + reviewer — who computed and who verified each statistic
- Dataset version/date — statistics are reproducible from the referenced dataset version
Statistical thresholds (what we will NOT publish)
- No rate (e.g. response rate) with fewer than 100 pitches sent in the pooled dataset — below that we publish counts and caveats, not percentages.
- No segment comparison (e.g. founders vs consultants) with fewer than 30 pitches per segment.
- No median time-to-response with fewer than 30 responses; we report median + IQR, not the mean.
- No predictive claims ("you should pitch X") — only descriptive findings.
Privacy
All benchmark statistics are computed from anonymized, aggregated event data. AstraMedia never stores or publishes message content, journalist personal information, email addresses, or customer-identifying details. No outcome is ever attributed to an individual customer without a documented written release.
What we will not claim
- No platform or agency can guarantee media coverage, and AstraMedia does not claim it can.
- We do not publish customer outcomes from internal access alone; outcomes require a customer's confirmed report or the product event model.
- We publish no star ratings, scores, or "X out of 10 users" aggregates.
- Every public claim on this page and in AstraMedia research is traceable to one of three sources: a named customer, a documented dataset, or a clearly identified company source.
Status
Methodology: final. This page is the methodology.
Statistics: pending data collection. The benchmark dataset publishes when the product instrumentation gate (outreach lifecycle events) is live and the thresholds above are met. Last reviewed: 2026-08-12.
Questions about this methodology?
Contact the AstraMedia team at contact@astramedia.app.