Methodology
What we measure, and what we cannot
This site publishes numbers. This page says where they come from, what is excluded, and which claims we are not in a position to make.
The corpus
Aggregate figures come from public Instagram data for accounts connected to PodSwap, read from public profiles only. There is no access to Instagram Insights, so reach and impressions are never reported — only engagement that is publicly countable.
Exclusions
- Accounts under 100 followers are excluded from bracket statistics — engagement rates at that size are too volatile to average meaningfully.
- Engagement rates above 50% are capped at 50% and labelled as capped, rather than dropped or silently rewritten. Niche leaders are capped at 75%.
- Where a monthly report omits a metric entirely, that section renders nothing. We do not substitute an estimate.
How delivery is confirmed
Delivery is confirmed by a combination of automated checks and pattern review across an account's whole history, rather than by any single test. We do not describe an engagement as verified unless it is, anywhere on this site.
Behind that sit two mechanisms that do not depend on taking anyone at their word: a 0–100 health score per account, which holds engagement for 24 hours below 95, and a dispute pipeline with screenshot proof, a right of reply and a human moderator. The penalties are symmetrical — a false report costs the person filing it exactly what failing to deliver costs the person reported.
We do not publish a breakdown of which checks apply where. A public map of where the checks are thinnest is an instruction manual for the small number of people who would use one.
Shadowban scanning
Instagram does not publish a shadowban status, so no tool can confirm one. Our scan is an inference from seven measurable signals in your own account history, and it reports a risk band rather than a verdict.
Availability
Exchanges open on a headcount rather than running continuously — see Live Rooms. Copy describing the exchange describes how it is designed to work, which is not the same as a promise that a given room is open at the moment you read it.
Pro and Max tiers appear in the product's data model but there is no billing integration. They are waitlist-only, and nothing on this site is for sale.
Corrections
If a figure here is wrong, we would rather know. The constants behind the matching engine, the Live Room thresholds and the Podscore rubric are all published on the features pages so they can be checked against what the product actually does.