How we test delivery speed

Definitions

Delivery speed is reported as two distinct numbers, not one. Conflating them is how cheaper panels make “instant delivery” claims that fall apart on bigger orders.

Time-to-first-delivery (TTFD)

Wall-clock seconds from the moment the order webhook fires to the moment the first follower / like / view lands on the target account.

Full-completion window

Wall-clock duration from order webhook to the final unit being delivered. We report the 90th percentile, not the average — averages hide the slow tail.

Test protocol

Each calendar month we place a fixed grid of 27 test orders on our own services using test accounts we own and operate. The grid is held constant so month-over-month numbers are directly comparable.

Platform

Instagram, TikTok, YouTube. Each platform has its own rate limits and delivery rails.

Service type

Followers, likes, views. Likes and views deliver faster than followers — separating them prevents one favourable mix from inflating the headline.

Order size

Small, medium, large. Small is the lowest tier; large is the largest tier we sell. Catches pacing bottlenecks the small tier hides.

3 platforms × 3 service types × 3 sizes = 27 orders per month. With a 30-day rolling window we always have one full month of data plus the partial current month feeding the dashboard.

Instrumentation

TTFD is measured directly from the same webhook events the order pipeline already emits for billing and refund logic — no separate stopwatch process. The pipeline records:

• order.received — webhook fires the moment the gateway acknowledges payment.
• order.first_delivery — fires on the first observed unit on the target account. Polling cadence is 30 seconds.
• order.completed — fires when the delivered count reaches the ordered amount.

The 30-second polling cadence puts a floor on TTFD precision: any TTFD below 30 seconds is reported as “<30s” rather than a specific number, because we cannot distinguish 5 seconds from 25 seconds with this instrument.

What “fast” means in numbers

Everything we publish on /buy-* is the median TTFD and the 90th-percentile completion time across the most recent 30 days of test orders. The targets we hold ourselves to:

Likes / views

< 5 minutes to first delivery·Within 6 hours to full completion (P90)

Followers

< 15 minutes to first delivery·Within 24 hours to full completion (P90)

Live current-month numbers are published on the status page and refresh on each completed test order.

Why we report P90, not the average

Order-completion time has a long right tail. A single stalled batch can be 100× the median while every other order in the same hour completes normally. Averaging that into the headline number produces a misleadingly slow figure — and reporting only the median hides the fact that a slow tail exists.

P90 means “9 out of 10 orders complete in this time or less.” It is the number a buyer should actually plan around.

Out of scope

• Specific competitor TTFDs are not tested here — see the compare reviews where each head-to-head includes a competitor TTFD measurement under the same protocol.
• Platform-side rate-limit incidents (a TikTok-side rollback or an Instagram-side content moderation pause) are excluded from headline numbers because they are not under our control. They appear as incidents on the status page instead.
• Custom-quoted enterprise orders use a separate pipeline with bespoke SLAs and are not part of the 30-day rolling window above.