Research Methodology
How MonkeyCam research studies are measured, cleaned, and published — in plain English.
All results published in MonkeyCam research describe only anonymous sessions observed on MonkeyCam. They do not automatically represent all random video chat platforms.
What Counts as a Match
A match is counted when two people are paired into the same private 1-on-1 video room. A match is recorded even if one side skips immediately — it simply means a pairing happened, not that a conversation did.
What Counts as Video Ready
Video is counted as ready when the first frame of the other person's video appears on screen (remote video first frame). This is the moment a match becomes a visible, face-to-face encounter rather than just a technical connection.
What Counts as a Valid Conversation
A conversation is counted as valid when a match survives beyond the initial skip window — meaning both people stayed long enough for a real exchange to be possible, rather than one side leaving within the first moments. Very short pairings are treated as skipped matches, not conversations.
How Skips Are Measured
A skip is recorded when a person ends a match and moves to the next one, whether by tapping skip or by leaving the room before a valid conversation begins. Skips are counted per session, so a study can report how many matches typically happen before a real conversation starts.
Anonymous Sessions
MonkeyCam does not require accounts. Sessions are identified by rotating anonymous ids that are not tied to an account, a name, or a persistent profile. Sessions can be chained within a browser tab, but the same person cannot be followed across tabs, devices, or visits.
Device Classification
Devices are classified from the browser's user agent (UA) string into broad classes such as mobile or desktop. No device fingerprints, precise hardware details, or other identifiers are used for research classification.
Geographic Limitations
Geography is recorded only at a broad country or region level. IP addresses are not stored for research purposes, and no fine-grained location (city, coordinates, or similar) is collected or published.
Sample Limitations
Every study sample consists only of people who used MonkeyCam during the study period. The sample is not a survey of the general population and not a panel of all random video chat users. Audience mix, time of day, and season can all influence the numbers, and a different period could produce different results.
Bot Filtering and Data Cleaning
Before analysis, events are validated and suspicious or malformed activity is flagged — for example automated clients, impossible event sequences, or sessions that never loaded video. Invalid events are flagged and excluded from the published metrics, not silently deleted, so the cleaning process itself stays auditable.
Study Periods and Snapshots
Each study runs for a defined period with a minimum sample size. When a study is published, its numbers are frozen as a snapshot of that period: the sample size, metrics, and findings on the page always refer to the stated study period and are not quietly rewritten afterwards.
Data Refresh Cadence
Studies collect data continuously and are published once their minimum sample and study-period requirements are met and the results pass review. A study page shows a collection-in-progress status until its first snapshot is published; after publication, new periods may be published as new snapshots.
Privacy
Research data contains no names, email addresses, phone numbers, chat content, or stored video. MonkeyCam research measures events — matches, skips, timing, device class — and nothing about who a person is or what was said or shown in a conversation.