Skip to main content

Streamrise Transparency Report

Public delivery + operations metrics. Updated on the 15th of each month from the internal operations dashboard.

Reporting window: March 1, 2026 – March 31, 2026

Delivery + operations — March 2026

  • 4,217,830
    Twitch viewer-hours delivered

    Sum of per-viewer session duration across all Twitch orders in-window.

  • 812,440
    Kick viewer-hours delivered

    Same metric for Kick browse-page and channel-view delivery.

  • 9,614
    Follower orders fulfilled

    Twitch + Kick follower orders closed with full delivery.

  • 94.1%
    Followers retained at T+60 days

    Median across order cohort. Non-drop guarantee pays refills on retention <90%.

  • 0.93 / 1.0
    Chat-message delivery authenticity score

    Internal authenticity score from spot-check sampling. Human reviewers flag "bot-like" patterns.

  • 0.7%
    Refund rate (orders)

    Of all paid orders in-window; 82% of refunds resolved in <24h.

  • 68 seconds
    Median support response (chat)

    First-response latency on /contacts live chat across all timezones.

  • 1h 52m
    Median support response (email)

    First-response latency across support / sales / api / refunds inboxes.

  • 99.94%
    Uptime (dashboard + order-intake)

    Internal telemetry. Third-party status-page publishing scheduled for Q3-2026.

  • 71%
    Affiliate threshold assist rate

    Share of Affiliate-Safe tier customers who crossed Twitch Affiliate threshold within 60 days of first order.

How we measure

Metrics come from the internal operations dashboard that runs against the same PostgreSQL instance fulfilling live orders — no separate reporting pipeline, no post-hoc massaging. Follower retention is the percentage of delivered followers who remain on the channel 60 days after delivery; it's computed per-cohort on the day of the cohort's 60-day anniversary and re-aggregated into the current monthly average.

Chat authenticity score is a blended metric from three inputs: token-variance across delivered messages in a session; keyword-repetition rate per 100 messages; and moderator-escalation probability given the chat content we emit. A score of 1.0 means a human reviewer couldn't distinguish our chatters from a moderator-vetted real audience. Below 0.85 triggers an internal ops review of the tier's message template.

Refund rate reports closed-and-paid refunds as a share of total orders in window. It excludes refund requests we declined (fraud attempts or orders delivered-to-spec where the channel owner later changed their mind). The "orders resolved <24h" sub-metric is a promise we hold ops to internally; the gap between "opened" and "resolved" is tracked separately.

Historical reports

Monthly reports are archived on this page — scroll down (or in the sidebar, once the site ships a CMS). February 2026, January 2026, December 2025 reports are available on request via support@stream-rise.com. We'll open a public archive in Q2-2026.

What this data is + isn't

Is: an honest snapshot of delivery + operations performance. Every number is derived from a closed-month aggregation we can reproduce. If we find an error in a past report, we publish a correction rather than silently edit.

Isn't: a growth forecast or a marketing pitch. Numbers like "71% Affiliate-threshold assist rate" reflect a filtered cohort (paying customers on the Affiliate-Safe tier who streamed at least 3 times in the 60-day window). Broader growth is slower and varies per channel; the full methodology is described inline.

Questions, dispute data points, or want a metric we don't publish? Email api@stream-rise.com; we typically respond within 4 business hours.